Empirical Correlation of Soil

Liquefaction Based on SPT TV-Value

and Fines Content Kohji Tokimatsu &
Yoshiaki Yoshimi
Visit to download the full and correct content document:
https://ebookmass.com/product/empirical-correlation-of-soil-liquefaction-based-on-spt
-tv-value-and-fines-content-kohji-tokimatsu-yoshiaki-yoshimi/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

The Effect of Fines on Critical State and Liquefaction

Resistance Characteristics of Non-Plastic Silty Sands
Anthi Papadopoulou & Theodora Tika

https://ebookmass.com/product/the-effect-of-fines-on-critical-
state-and-liquefaction-resistance-characteristics-of-non-plastic-
silty-sands-anthi-papadopoulou-theodora-tika/

SPT-based probabilistic and deterministic assessment of

seismic soil liquefaction triggering hazard K. Onder
Cetin & Raymond B. Seed & Robert E. Kayen & Robb E.S.
Moss & H. Tolga Bilge & Makbule Ilgac & Khaled
Chowdhury
https://ebookmass.com/product/spt-based-probabilistic-and-
deterministic-assessment-of-seismic-soil-liquefaction-triggering-
hazard-k-onder-cetin-raymond-b-seed-robert-e-kayen-robb-e-s-moss-
h-tolga-bilge-makbule-ilgac/

Effects of Sand Compaction on Liquefaction During the

Tokachioki Earthquake Yorihiko Ohsaki

https://ebookmass.com/product/effects-of-sand-compaction-on-
liquefaction-during-the-tokachioki-earthquake-yorihiko-ohsaki/

Chemical, Gas, and Biosensors for Internet of Things

and Related Applications Kohji Mitsubayashi

https://ebookmass.com/product/chemical-gas-and-biosensors-for-
internet-of-things-and-related-applications-kohji-mitsubayashi/
Standard Penetration Tests and Soil Liquefaction
Potential Evaluation Fumio Tatsuoka & Toshio Iwasaki &
Ken-Ichi Tokida & Susumu Yasuda & Makoto Hirose &
Tsuneo Imai & Masashi Kon-No
https://ebookmass.com/product/standard-penetration-tests-and-
soil-liquefaction-potential-evaluation-fumio-tatsuoka-toshio-
iwasaki-ken-ichi-tokida-susumu-yasuda-makoto-hirose-tsuneo-imai-
masashi-kon-no/

Naming Africans: On the Epistemic Value of Names

Oyèrónk■■ Oy■wùmí

https://ebookmass.com/product/naming-africans-on-the-epistemic-
value-of-names-oyeronke-oyewumi/

Andy Burnett On Trial 1958 TV Series Big Little Book

Big Little Books

https://ebookmass.com/product/andy-burnett-on-trial-1958-tv-
series-big-little-book-big-little-books/

Elsevier Weekblad - Week 26 - 2022 Gebruiker

https://ebookmass.com/product/elsevier-weekblad-
week-26-2022-gebruiker/

Fundamentals of Soil Ecology 3rd Edition

https://ebookmass.com/product/fundamentals-of-soil-ecology-3rd-
edition/
 SOILS AND FOUNDATIONS Vol.23, No.4, Dec. 1983
Japanese Society of Soil Mechanics and Foundation Engineering

EMPIRICAL CORRELATION OF SOIL LIQUEFACTION

BASED ON SPT N-VALUE AND FINES CONTENT

KoHJI ToKIMATsu* and Y OSHIAKI Y OSHIMI**

ABSTRACT
A critical review of field performance of sandy soil deposits during past earthquakes is
conducted with special emphasis being placed on Standard Penetration Test N-values and
fines content. The field relationship between adjusted dynamic shear stress ratio and normal-
ized SPT N-values together with laboratory tests on undisturbed sands indicate that (1)
sands containing more than 10 percent fines has much greater resistance to liquefaction than
clean sands having the same SPT N-values, (2) extensive damage would not occur for clean
sands with SPT Ncvalues (N-values normalized for effective overburden stress of 1 kgf/cm 2)
greater than 25, silty sands containing more than 10 percent fines with SPT N 1-values
greater than 20, or sandy silts with more than 20 percent clay, and (3) sands containing
gravel particles seem to have less resistance to liquefaction than clean sands without gravel
having the same SPT N-values.
On the basis of the above findings, an improved empirical chart separating liquefiable and
non-liquefiable conditions is presented in terms of dynamic shear stress ratio, SPT N-values,
fines content, and shear strain amplitude.

Key words: case history, earthquake, earthquake damage, l~guefaction, penet~~tion___!~~'

sample disturbance, sand, sandy soil (IGC: D 7/C 3)

primary, if not the only, justification for

INTRODUCTION
Despite its limitation as to lack of rigid
standardization, the standard penetration
test (SPT) N-value has been employed as a
convenient parameter to express resistance of
sandy soils to liquefaction. When the concept
of critical N-value was first proposed by
Koizumi (1966) and Kishida (1966) based on
their observations of the Niigata earthquake
of 1964, the availability of data was the
using the SPT results to express resistance
of sands to liquefaction. For several years
afterwards, the emphasis was shifted toward
the use of relative density that could readily
be correlated with laboratory test results
(Seed and Idriss, 1971).
More recently, however, the SPT has
received renewed attention as various factors
other than the relative density have been
recognized to have significant influences on

* Assistant, Tokyo Institute of Technology, Meguro-ku, Tokyo (currently on leave as Visiting

Scholar at University of California, Berkeley).
** Professor, Tokyo Institute of Technology, Meguro-ku, Tokyo.
Manuscript was received for review on March 31, 1983.
Written discussions on this paper should be submitted before October 1, 1984.

This is an Open Access article under the CC-BY-NC-ND license.

NII-Electronic Library Service
 EMPIRICAL CORRELATION OF LIQUEFACTION 57

the liquefaction resistance of sands. The Pier No. 7

( 6•13cm)
Pier No. 12
( S•IOcm)
Pier No.2!
(8•&9em)
Pier No.30
(8•!50cml
Pier No.49
( 6• 5em l
SPT N-volue SPT N-volut SPT N-value SPT N-volue SPT N-volue
following conditions seem to support the use 0 f02040 f02040 f 0 20 40 f 0 20 40 f 0 20 40

of the SPT N-value to express directly the

resistance of sands to liquefaction, if a more
standardized SPT procedure is used or an
appropriate correction is made for adjusting
energy loss due to friction between cathead
and rope:
1. The SPT is an in situ test which
reflects stress history and strain history
effects, soil fabric, and horizontal effective
Legend: S • Sand , C • Cloy , MS • Medium Sand , FS • Fine Sand • G • Gravtl
stress, in addition to the combined effect of So • Surface Soil, f • Fints Content in ptrctnt

the relative density and the vertical stress Fig. 1. Results of soil investigations
(Seed, 1979). All of the above factors are conducted in 1964 along Arakawa
Bridge of the Tohoku line with lJ
known to influence the resistance of sands
indicating settlement of spread foot-
to liquefaction but are difficult to retain in
ing due to the Kanto Earthquake of
most so-called "undisturbed" samples. 1923 (after Kodera, 1964)
2. Numerous case histories of soil lique-
faction during past earthquakes are available
for which the SPT N-values before the
earthquakes are known. The method based
on field performance with these SPT N-
values can, therefore, reflect in situ soil
characteristics under real stress conditions
during earthquakes which are difficult to be 10

thoroughly simulated in the laboratory. Heavy Damage Sife

Besides, abundant SPT data in potentially Pier No.

-21
---- 30
liquefiable soils will become useful for future c
<1> 50
0 Depth in meter
earthquakes. e
~

3. The SPT is primarily a shear strength

test under essentially undrained conditions o~~~~~~~L-~~~~~~
0.001 0.01 0.1 10

because of rapid strain rate (Schmertmann, Grain Size (mml

1977). Fig. 2. Grain size distribution curves for

4. The SPT yields representative samples the soil shown in Fig. 1 (after Kodera,
of soil whose index properties can be deter- 1954)
mined. This is a clear advantage of the SPT
over the cone penetration test alone. o.-.---.-.-P.4-9~~~,.~-. 0.15 O.l
- P.49
5. The low cost of the SPT enables one P.l2~6 0
~2 P.I2
P.7 1 ~~ o.r o.23P. 7
to carry out many tests so as to reduce the tO 20

risk of missing weak spots. c

<1>

~ 40
On the other hand, it has often been a; ..i.P.30 P. 30 °ai6
(f)
pointed out that the presence of fines tend to E 60
8
P.21 P.21 0~.28
:::>
reduce the SPT N-value of sands without E
·;;;

affecting their shear strength so much. ~ 80 11...(a l 1..::{::-b-)---:-L---::'::"-~-'-::-'

--:-2---:'5~.,-~::1o--=z'::-o----:5:':0---' o.o5 0.1 o.z 0.5
Kodera (1964) reported that some piers of Fines Conlenl {%) Mean Grain Size, 0 50 { mm)

the Old Arakawa Bridge of the Tohoku line Fig. 3. Effects of grain size distribution
across the Ara river on clean sands settled on settlement of bridge piers during
during the Kanto Earthquake of 1923 con- the Kanto Earthquake of 1923 (data
siderably more than those on silty sands, from Kodera, 1964)

NII-Electronic Library Service

 58 TOKIMATSU AND YOSHIMI
despite the fact that the SPT N-values for
the latter were smaller "probably because the
clean sands were sensitive to vibration. "
The results of soil investigations conducted
in 1963 a~e shown in Figs. 1 and 2, from
which Fig. 3 has been prepared by the au-
thors. It should be noted that the soil data
were obtained about 40 years after the
earthquake. Nevertheless, the grain size
distribution has not probably changed al-
though the SPT N-values may have changed
where the settlement was large. The sig-
nificant settlement due to the earthquake and
the soil types seem to indicate that the soil
was probably liquefied. Fig. 3(a) appears to
support Kodera's observations, and shows
rather clearly that the cleaner sands contain-
ing less than eight percent fines suffered
much larger settlement than the silty sands
with more than 20 percent fines, whereas
no such clear separation exists in terms of
the mean grain size as shown in Fig. 3(b).
Using the fines content as an index pa-
rameter for estimating the liquefaction re-
sistance has the following advantages:
1. The fines content is better correlated
than the mean grain size with the degree
of damage due to soil liquefaction as shown
in Fig. 3.
2. The fines content is probably better
related with soil consistency which in turn
is related to undisturbed shear strength of
soil.
3. The fines content can be determined
more easily than the mean grain size by
washing a soil sample through a 74 ,um sieve,
e. g., by "the method of test for amount of
material in soils finer than the 74 ,um sieve"
proposed by the Japanese Society of Soil
Mechanics and Foundation Engineering
(Designation JSF T 22-1971).
On the basis of the above discussion, a
critical review of field performance of sandy
soil deposits during past earthquakes is con-
ductecl in order to establish a more reliable
empirical chart for estimating their lique-
faction resistance.
DETERMINATION OF FACTORS FOR
LIQUEFACTION CRITERIA
In order to review case histories of soil
liquefaction, a special attempt has been
made: (1) to reexamine shear stress devel-
oped during earthquakes, (2) to assess the
effect of the procedure of the SPT and soil
type on the liquefaction resistance, and (3)
to classify the degree of liquefaction.
Shear Stress Ratio to Represent Seismic
Ground Motions
Based on extensive laboratory test results
of liquefaction of saturated sands, effects of
seismic ground motions causing liquefaction
may be represented by two quantities: hori-
zontal ground acceleration and number of
cycles of significant ground motions. The
finding is incorporated in the following
equation for dynamic shear stress ratio for a
given depth at a given site.
--r-
d
- -a-max
---do
rr ( 1)
Oo' - g oo' d n
in which -r a =amplitude of uniform shear
stress cycles equivalent to actual seismic
shear stress time history, amax=the maxi-
mum horizontal acceleration at ground sur-
face, rJ 0 1 =initial effective vertical stress,
o 0 =initial vertical stress contributing to the
shear stress defined by
Oo= lz TtdZ ( 2)
where rt=unit weight of soil and z=depth
below the ground surface, and r a and r n
are correction factors in terms of depth,
z in meters, and earthquake magnitude, M,
respectively, as follows:
ra=1-0.015z (Iwasaki et al, 1978b)
( 3)
rn=0.1(M-1) (4)
Note that o 0 cannot be equivalent to the
total vertical stress if water table is located
above the ground surface, because water
above ground surface cannot transmit shear
stresses.
In order to facilitate comparison of field
NII-Electronic Library Service

Table 1. Relationship among
earthquake magnitude,
number of cycles, and rn
l':arthquat:e Number
11agnitude of Cycles
5.5
6.5 6 0.54
7 10
7.5 15
8.3 25
I.O,----.-------r----,----,
- rn
0 Data from Table I
6 7
Ecr~hquake Magnitude,
Fig. 4. Relationship between rn and
earthquake magnitude
behavior during earthquakes of different
magnitudes, the factor r n in Eq. ( 4) is intro-
duced so that a given number of cycles,
Nz, of 0. 65 times the maxim urn shear stress
amplitude will be converted into 15 cycles
of rn times the maximum shear stress am-
plitude, on the basis of the relationship
between number of cycles of 0. 65 times the
maximum shear stress and earthquake mag-
nitude (Seed et al, 1975) as shown in Table
1, and the assumption that the relationship
between dynamic shear stress and number
of cycles on a log-log graph is a straight
line having a slope of 0. 2 as defined by
rn=O. 65(Nz/15) 0 • 2
The exact relationship between r n and M
based on Eq. (5) is shown in Table 1 which
may well be represented by Eq. ( 4) as shown
in Fig. 4. It can be noted that if M=7. 5,
Eq. (4) gives rn=O. 65 which is familiar in
currently available simplified procedures.
\Vhen the expressions for r d and r n are
incorporated, Eq. (1) becomes
-rd
(f o'
= 0. 1 (M -1)
g
EMPIRICAL CORRELATION OF LIQUEFACTION 59
The maximum acceleration in Eqs. (1) and
(6) has been estimated on the basis of various
field observations including those of over-
r turned tombstones and collapse ratio of
n
wooden houses as well as empirical relation-
0.47
ships between earthquake magnitude and
epicentral distance and, if any, strong motion
0.60
records registered near the site. In order to
0.65
estimate the maximum horizontal accelera-
0. 72
tion from overturned tombstone~, the follow-
ing equation 1s introduced based on static
equilibrium.
1Xmax K
-~-~
(7)
g 1+tcK
in \vhich K is the resultant seismic coeffi-
cient defined by the critical width to height
=O.I(M-1) ratio, B/H, of a tombstone, which would
just overturn during an earthquake, and tc
8
denotes the ratio of the maximum vertical
M
acceleration to the maximum horizontal
acceleration. In the following, the coefficient
tc is rather arbitrarily assigned a value of
0. 5 for determining a max/ g, because many
studies have shown that it is within a range
between 1/2 and 2/3.
SPT N-value to Express Liquefaction
Resistance
Recent studies by Schmertmann (1977) and
Kovacs et al (1977) have shown that poten-
tial variability may exsist in the SPT N-
value despite the adjective "standard."
Among various factors, the amount of en-
ergy delivered into sampler rods appears to be
the most important factor affecting SPT
N-values. The SPT N-values used in the
following pages are mainly obtained in Japan
( 5) and, therefore, measured according to the
Method of Penetration Test for Soils, JIS
A 1219-1961, that requires a free fall of the
hammer.
On the basis of a questionnaire survey
conducted recently, Yoshimi and Tokimatsu
(1983) showed that the following three proce-
dures to drop the hammer onto the rod
during the SPT measurement are frequently
4ao (1-0. 015 z)
used in current Japanese practice.
( 1 ) The trip monkey (tonbi) method
(6) ( 2) The cathead (pulley) and rope meth-
NII-Electronic Library Service
 60 TOKIMA TSU AND YOSHIMI

special caution would produce the same values

0 A Data from Yoshimi and Tokimotsu (1983)
as those by the trip monkey method. Eq. (8)
o A Dota from Tokimot&u and Yoshimi (1981)
40
~ L--~-= ~~~~ Sand is consistent with the findings by Nishizawa
Q;
., :::!! 30
et al (1980) in which the ratio of the net
::>
g~
>.
energy delivered to the rod to the theoretical
' 0
z:::!! 20 maximum is 0. 63 to 0. 72 for the cathead
1--a.
0.. •t:
(J)I--
., and rope method and 0. 8 to 0. 9 for the free
= 10 fall type method.
Since the ratio of the net energy in the
OQ I0 20 30 40 50
U.S. A. was estimated to be about 0. 5 to
SPT N-value
by the Cat head and Rape Methad 0. 6 by Kovacs and Salomone (1982), the fol-
Fig. 5. Effect of procedure of lowing relation is assumed for interpreting
SPT on measured SPT N- SPT N-values from foreign countries.
values in Japan Ncf = 1. 4 NtJ (9)
in which cf means the cathead and rope
od with two turns of rope around the
method in foreign countries.
cat head
( 3 ) The same method as (2) except for
Normalized SPT N-value
a manner of releasing the rope, in that the
Since it is well known that the SPT
rope is completely thrown off the cathead in
N-values are influenced by the effective
order to reduce rope friction.
confining pressure in the soil as well as the
In methods (1) and (3) one intends to pro-
soil density which may reflect the undrained
duce a free fall condition, while in method
strength of the soil, corrected SPT N-values
(2) one does not attempt to reduce the
for a reference confining pressure have oc-
friction.
casionally been used for practical purposes
In order to clarify the influence of energy
(Seed, 1979). The SPT N-values normalized
loss due to the friction on measured SPT
for a 0 '=1kgf/cm 2 (98kPa), N 1 , which is
N-values, comparative field tests for clean
adopted hereafter, may be approximately
sands were conducted by Yoshimi and Toki-
g1ven by
matsu (1983) by the methods (1) and (2).
Fig. 5 shows the fmal results of the field N 1 =C N=_l.J---N (10)
N 0" 0 '+0.7
tests. The results are in fairly good agree-
ment with the previous findings by Toki- m which C N is a function of the effective
matsu and Yoshimi (1981) of which data vertical stress, a 0 ', in kgf/cm 2 , at the time
are also replotted in the figure by using the
mean values. Since all data statistically 0 0.5 1.5 2.0
show that the SPT N-value by the cathead -
NE
0

1.7
and rope method is greater by about 20 ~
~
CN= (J~{kgf/cm2 !+0.7

after Seed ( 1979) ;-'

percent than the trip monkey method for '0
b .,.."'""'"
any N-values up to 40, the following equa- ., 0.5

tion could, therefore, be used to correlate "'

i'!
(;) ,///
the relation between the two methods in
Japan.
( 8) o, = 40 to 60 % j
in which cj means the cathead and rope ,'f-Dr =60 to 80%

method in Japan, and tj means the trip ,l!

J.5L----.t___-..JJ..i._J.__ _ _J . __ __ _ i

monkey method in Japan. Fig. 6. Correction factor CN for SPT N-

It is conceivable, on the other hand, that values in terms of effective vertical
N-values measured by the method (3) with stress

NII-Electronic Library Service

 EMPIRICAL CORRELATION OF LIQUEFACTION
when and at the depth where the penetration
test was conducted. Eq. (10) is based on a
simplified relation by Meyerhof (1957) which
in turn was based on the test results by
Gibbs and Holtz (1957) for the effective
vertical stress up to 2. 8 kgf/cm 2 (275 kPa),
and is also equivalent to that recently pro-
posed by Seed (1979) as shown in Fig. 6 at
least for o 0 '~1.5kgf/cm 2 (147kPa) which
may cover the range of general interest for
soil liquefaction. Although the equation
would produce a conservative estimation in
SPT N 1-values for o 0 '~1. 5 kgf/cm 2 (147 kPa)
in comparison with the value by Seed (1979),
this tendency seems favorable considering the
decrease in the net energy transmitted to the
rod longer than 20m (Nishizawa, 1980), as
well as the reduction in liquefaction resist-
ance due to changes in dilatancy with in·
creasing confining pressures.
Classification of Degree of Liquefaction
Sand boils where no artesian conditions or
broken water lines exist are usually con-
sidered as a sign of liquefaction. Excessive
settlement of either ground surface or heavy
structures may constitute more positive
evidence of liquefaction. In this paper, the
degree of liquefaction is classified in to four
classes as shown in Table 2: extensive
liquefaction, moderate liquefaction, marginal
condition, and no liquefaction.
Field behavior in which either ground sur-
face settled more than about two percent of
the thickness of the sand layer which might
have liquefied or heavy structures settled
more than 20 em is classified as extensive
liquefaction. A site where a sign of lique-
Table 2. Degree of liquefaction
Degree of
Liquefaction Criteria based on field observation
In addition to sand boils, ground
Extensive
Liquefaction
settles more than 2
thickness, or heavy
settle more than 20
% of stratum
structures
em.
In addition to sand boils, ground
Moderate
Liquefaction
settles less than 2
thickness, or heavy
settle less than 20
% of stratum
structures
em.
Marginal Site separating liquefied and
Site non-liquefied area
No Neither sand boils nor settlement
Liquefaction are observed
61
faction was apparent but settlement and
damage were estimated to have been less than
the values cited above is called moderate
liquefaction. A site where neither sand boil
nor settlement was observed is classified as
no liquefaction, while a site located on
the boundary separating liquefaction and
no liquefaction conditions is considered to
be marginal.
FIELD CORRELATION OF SOIL LIQUE-
FACTION BASED ON SPT N-VALUE
AND FINES CONTENT
Compilation of Data
Many investigators have reported field
evidence of soil liquefaction during strong
earthquakes of which more than 70 case
histories in Japan during 10 earthquakes as
well as about 20 supplemental data outside
Japan are available as shown in Table 3.
SPT N-values in Japan disclosed to be
measured by the method mentioned previ-
ously are indicated in the table, but the
others are, hereinafter, considered to be
determined by a free fall type method be-
cause the questionnaire survey by the authors
indicates that more than a half of SPT
measurements in Japan are conducted in an
attempt to produce a free fall of the hammer
and this assumption is on the safe side.
Fig. 7 shows the liquefied sites during the
Mino-Owari and the Tohnankai Earthquakes
as well as the collapse ratios of wooden
houses and contours of the width to height
ratios of overturned tombstones during the
Mino-Owari Earthquake. Note that the
collapse ratio during the Mino-Owari Earth-
quake is as much as 90 percent which might
have corresponded to a horizontal acceleration
Symbol
in Figures of 0. 3 g. The SPT N-values for these sites
were determined by the trip monkey meth-
• od .
One of the most catastrophic failures
occurred during the Fukui Earthquake of
• 1948, resulting in almost 100 percent collapse
ratio of wooden houses near the fault as
D.
shown in Fig. 8. The width to height ratios
0
of overturned tombstones around Maruoka
NII-Electronic Library Service
 62 TOKIMATSU AND YOSHIMI

Table 3. Soil and seismic data

Fault Line~- ·\,.Epicenter Mino-Owari Earthquake
···~.
{Nobi) of 10/28, 1891

Approximate Area of
Energ\eose

( ?D%) Ratio dunng

No01 Earthquake '",

/,>' ( 1:::!.'~:1 )· -:CoiL:J~sa Ratio during >

l the Tih:1ar:.1.ot Earthquake'"
cft$r lid,:' 1'977,197~)
/'fohnankai EorthqJJake
•• _ / / of 1217, 1944
.•• M; 8.0 "======~50 km
/Epicenter 1;~:~: ~
8

Fig. 7. BjH ratios of overturned

tombstones and collapse ratios
of wooden houses during the
Mino-Owari Earthquake and
the Tohnankai Earthquake

i0.2(16 0
10.114
0.3501 '5
0.292 4
o.nt.. 60'
o.z4sj o'
~:;~;,1
0.219
0.281
XI
7
12
0. ~~
1.6
1.2
'•.4
8
25
10.241 17 0.15
0.172 5 0 0.14 2.1 I
0.214 4 2.6
0.179 5
o··
0.53
T.36 2.5 Iwaoaki et a1,1J
0.223 0.41 2.5 1981
0.202 0 0.3 3.5 Tsuchida et al,
0.180 10 4 0 0.25 3.5 1979 1

5 0 0.35 2.5 !

20 0 0.15 15
26 0 0.12
0.202 3 2 0.35 2.3
4 0 0.30 5''
I 0.209 11
0.30' s'J S~.:rfa-:a Fou!ri11g
0.188 12 3 0
0.183
(<.195110
10 0.15 2.0
0.18 2.4
Tshihara£>tal,
1980 / Fukui Earthquake
of 6/28 '48
•
..
'lzu 10.244 90 0 0 eta1,
!tOl/14/1978) I M = 7.3
Clljbakcnchubu 6.1
1(09/25/1')80)
lo.os:,
,0.143
0 0.18
0.18
0 36.1 • N
0.17 0
136.2° E

-----,---~0.07 0 19¥$ et a1,

I'·' • Ross, 1968

I ~: ;~:
•• I R"s, 1968
lo.2"
Snnd&
~~rdVCl
I
••
4.5
:o.l'' 1
•• lteeetal,1975 Tckcya 2 /

I
6.5 lo.o7 • .£;;;~ !:7:1,

4.6
4,6
I
1.0
1.0 •
A
Seed et al,
1981
4 3.3 7.0 1.0 0
75'%
Tan shan
(07/28/1976)
7.8 Lutai 3
Lutai 51 0.5 11 2.2 1.3 7 31 0.2€) 0.192
so'~~
50'1) 15
s 0.07
0.07
6''
20~' •• Zhou, 1981
lOkm \ 0.50 8/H ratio
Lutai 52 o.s 11 2.2 1.3 all o.zo o.l92 so''115 0.07 20 .., A =======I
Imperial 6.6 Heber Rd 1 1.8 0.74 0.52 3lll 43.2 0.60 0,449 11 0.12 2.3 0 Benett ct al,

.
Valley Heber Rd 4 1.8 0.74 0.52 4" 5;6 0.60 0.449 25 5 0.12 11 • 1981
(10/15/1979) Heber Rd 7 1.8 0.74 0.52 11 11 15.3 0.60 0.449 19 1 0.10 2.4 0
River Park6 0.2
River Park6 0.2
0.38
0.95
0.20
0.47
3"
7''
5.7
10.1
0.20
0..20
0.206
0.209
66
34
15
4
0.04
0.09
• Fig. 8. BjH ratios of overturned
tombstones and collapse ratios
tq:;cnd: FC =Fines Content, CC"' Clay Content, GC =Gravel Content, Uc =Coefficient of Uniformity
e"' Extensive Liquefaction, 8 • Moder<'!te Liquefaction, 6 = ~larginal Site~ 0"' No Liquefaction
1) =determined ·by a free fall type method, 2) =corrected by Eq, (9)
of wooden houses during the
3) = estimated from Cone Penetration Test CJc-value, 4) mean value, 5) estimated from other data
6) "" corrt.'ctcd in terms of stress, bec.Juse c:~ at the Sl'T measurement \-.'as not the same as tho.t during the eartl1quake Fukui Earthquake

NII-Electronic Library Service

 EMPIRICAL CORRELATION OF LIQUEFACTION
;~~ri.~u':heH2~~R~+:c~~,~~~onker
B ·.··~Maximum Horizontal Acceleration
l dunng the 6/12 '78 Earthquake
( } ; Maximum Horizontal Acceleration
:Estimated from B/H ratios \..,
[ during the 6/ 12 '78 Eorthqu~~
Arahom··
(0.201
Fig. 9. Maximum horizontal ground accel-
erations and B/H ratios of overturned
tombstones during the Miyagiken-oki
Earthquakes of 1978
Town exceeds 0. 5, which leads to a hori-
zontal acceleration of about 0. 4 g or more.
The SPT N-values of the sites were deter-
mined by the trip monkey method.
A great amount of data is supplemented
by extensive field investigations conducted
before and after two Miyagiken-Oki Earth-
quakes that took place with a few months
interval in 1978, i.e., February 20 earthquake
of magnitude 6. 7 and June 12 earthquake of
magnitude 7. 4. The maximum horizontal
accelerations recorded during the earthquakes
and the width to height ratios of overturned
tombstones near the sites for the earthquake
of June 12 are shown in Fig. 9.
Field surveys after the Kanto Earthquake
of 1923 show a collapse ratio of wooden
houses of three percent and a B/ H ratio of
about 0. 25 near the Old Arakawa Bridge,
which leads to a horizontal acceleration of
about 0. 2 g.
Zhou (1981) presented field case histories
during the Tanshan Earthquake in which,
unfortunately, neither ground acceleration
nor SPT N-value was reported. Neverthe-
less, the facts that the earthquake intensities
of VII, VIII, and IX in Mainland China may
approximately correspond to the maximum
horizontal accelerations of 0. 1 g, 0. 2 g, and
0. 4 g, respectively, and that the relationship
between the Cone Penetration Test qc-value
and the SPT N-value in Mainland China is
63
M~6.7
2/20 '78
0
313°48' N
142°04'E
oo
:-.. Q> Qlooo
§6&18:>30'L4 I d 0
05 w 40 60 80 100
Silt Content (%)
Fig. 10. Triangular classification chart
for liquefied soils. The numerals
near the lower left corner indicate
the number of data points
given by qc=4. 5 N(Zhou, ·1981), as well as
the epicentral distance of the site enable one
to con vert the original data.
Because data concerning the other earth-
quakes are readily available elsewhere (see
references in Table 3), a review of estimat-
ing the maximum horizontal accelerations is
omitted here.
Triangular Classification Chart
A triangular classification chart for the
liquefied soil data in Table 3 is shown in
Fig. 10 in order to identify soil type which
may have liquefied. It is interesting to note
that more than a half of all liquefied data
fall within a range of fines content less than
five percent and that none of the soils con-
taining more than 20 percent clay has suffered
serious strength loss due to liquefaction.
Particularly remarkable is the fact that the
latter evidence is in good agreement with a
study in China (Seed and Idriss, 1981). The
data presented here, however, are probably
the first comprehensive evidence to clarify
that point.
It is also noted that soils with more than
60 percent fines, many of which must con-
tain more than 20 percent clay, hardly
liquefied except for a few cases including
mine tailings reported by Okusa et al (1980)
that did have no plasticity even though the
fines content was. as much as 90 percent.
This trend was alsp._ confirmed by Lee and
Fitton (1968) whose two silty sands (F, H)
with more than 50 percent silt but with a
plasticity index less than three percent
NII-Electronic Library Service
 64 TOKIMATSU AND YOSHIMI

showed low liquefaction resistance compared Table 4. Mean SPT Ncvalues,

with a sandy silt (G) which had a plasticity
index of 17. According to a study in Main-
land China (Finn, 1982), plasticity index of
10 seems to be the threshold for liquefaction.
Even though more field evidence is needed
to clarify this point, it is conceivable that
the plasticity index could be a promising
index for estimating liquefaction resistance
of silty soils, considering the fact that it is
more reliable yet easier to determine than
clay content.
Relationship Between Stress Ratio and SPT
Ncvalue
Figs. 11 and 12 show the relationship be-
tween either fines content or mean grain
size and SPT N 1-value for the liquefied soils.
There is a fairly well-defined trend in which
100 ,----,----.-----..,.---.---,
0 Extensive Liquefaction 0
80 \
~ I
q
6Q 0 I
c td
0> I
C I
8 40 0>
N 1, for liquefied soils
Fines Content N1 with more
(%) than 5% gravel
0 - 5 12 16
10 - 20 7
20 - 60 5
60 -
the SPT N 1-values for the liquefied soils
decrease with either an increase in fines
content or a decrease in mean grain size.
From Fig. 11, the mean N 1-values for
liquefied soils in terms of fines content are
shown in Table 4. Particularly noted. is a
significant reduction in SPT N 1-value for
silty sands with more than 10 percent fines
compared with clean sands, which would appear
to confirm the fact that the presence of
fines reduces SPT N-values without signifi-
e
cant changes in liquefaction resistance.
Moderate Liquefaction o • Although data with more than 60 percent
with more then ~Yo Qravel __j
fines are scarce, the value in Table 4 could
be reasonable and yet conservative consider-
ing a general tendency in which SPT N-
val ue decreases with an increase in fines
\

tl)
I
~ 0
0
\
content and field evidence in terms of fines
~ 9>r:Jo \
G: 20 \ (j o\, content shown in Fig. 10.
E:l\&o••'',, It is also noted in Fig. 11 and Table 4
\~~~J~~ that the mean SPT Ncvalue, N 1, for soils
0o ·,o· · 20 30 40 50
Normalized SPT N-value, N1 containing more than five percent gravel is
Fig. 11. Relationship between fines con- greater by roughly four than that without
tent and normalized SPT Ncvalue for gravel. This fact probably reflects an increase
liquefied soils in penetration resistance during SPT meas-
urements due to the presence of the large
1.4 particles.
Extensive Liquefaction 0 e

~ 1.2
•Moderate Liquefaction o • Although it may be seen in Fig. 12 that
wfth more than 5% !Jravel ---..1'
D 50 = 0. 2 mm could be used as an index to
J I,Q I 0 I
separate soil type, the mean SPT N-values
I
I I

i o.a : •: in terms of D 50 such as those defined in

(j) :.
,!:; 0.6
I
I 0 8
• I
:
Table 4 can hardly be estimated from this
e
<.!) :
I
0
0
.,
:
figure.
c 0.4 ap 'I 0 :
On the basis of the above findings, field
~ a ~) ~ ~<:_Q_J
0.2~·
.. ·: 0(
correlations between shear stress ratio and
. ·. e-"
0
0·-' normalized SPT N 1-value are shown in Figs.
o - 10 20 30 40 50
Normalized SPT N-value, N1 13 and 14 for sands with less than five
Fig. 12. Relationship between mean grain percent fines and for silty sands with more
size and normalized SPT N 1 -value for than 10 percent fines, respectively. The
liquefied soils numbers beside the data points indicate gravel

NII-Electronic Library Service

 EMPIRICAL CORRELATION OF LIQUEFACTION 65

0.7 r----r---,..----r---,..-----.---, not take place for silty sands with SPT
• Extensive Liquefaction
• Moderate Liquefaction N 1-values greater than about 15 which con-
" Marginal Condition
0.6 o No Liquefaction tain more than 10 percent fines. It may
Number· beside data point
Indicates gravel content also be noted that liquefied clean sands with
~ 0.5 SPT N 1-value around 20 tend to contain
'~ gravel particles.
0.4 0 Kishida (1969) pointed out that the coeffi-
• •'
.Q •" cient of uniformity might also play an
~ 0.3 • 02
important role controlling liquefaction re-
"'"' 0
sistance, i. e., the greater the coefficient of
~ 0.2
(/) uniformity, the higher the liquefaction
0
5 resistance becomes. Besides, on the basis of
Q)

65 0.1 field performace during the Niigata Earth-

Fines Content :;; 5%
quake, Ohsaki (1970) tentatively stated that
10 20 30 40 50 sands whose fines content is less than 10 or
Normalized SPT N-value, N 1 whose coefficient of uniformity is less than
Fig. 13 Field correlation between shear five are likely to liquefy. As far as the sandy
stress ratio and normalized standard soils are concerned, the coefficient of uni-
penetration test N 1-value for clean formity tends to become more than five when
sands the fines content exceeds 10 percent as shown
in Fig. 15 which was prepared from Table 3.
0.7 .-----...--~---.,---..----,---., The field evidence presented here seems,
• Extensive Liquefaction
• Moderate Liquefaction
" Marginal Condition
therefore, to be consistent with the early
0. 6 o No Liquefaction
Number beside data point
findings in the field.
indicates fines content
The boundaries separating liquefaction and
no liquefaction conditions for lower N-values
o"
seem to be much better defined in Figs. 13
0.4 o 35
and 14 than the case in which the data
points were classified on the basis of the
mean grain size as reported previously
(Tokimatsu and Yoshimi, 1981). This is due
probably to the careful review concerning
the shear stress and the SPT N-values as
well as using the fines content in place of
Fines Content~ 10%
the mean grain size.

Normalized SPT N -value, N1

Fig. 14. Field correlation between shear

stress ratio and normalized standard
penetration test Ncvalue for sands
containing more than 10% fines

content in Fig. 13 and fines content in Fig. 14.

It is apparent from these figures that:
(1) sands containing more than 10 percent
fines have much greater resistance than clean I0 I0 20 30 40 50 60 70
sands which have the same SPT Ncvalues, Fines Content (%)

(2) extensive damage would not occur for Fig. 15. Relationship between fines con-
clean sands with SPT Ncvalues greater than tent and coefficient of uniformity
about 20, and (3) extensive damage would for liquefied soils

NII-Electronic Library Service

 '66 TOKIMATSU AND YOSHIMI

COMPARISON WITH LABORATORY Double Amplitude Sampling Method

Axial Strain on which data
'TEST RESULTS 2.5'"/o 5% 7.5% ore based

1.0 A Cl 0 !n S1tu Freezing

Data from Laboratory Tests on Undisturbed A. • e Conventional
-----·-------
Sands
Since it is still uncertain whether the lack N = 15 cycles
0
·of data for dense sands or positive dilatancy :g 0.5
a::
·characteristics of dense sands has attributed
to the fact that no liquefaction did occur
beyond the critical N 1 values, which are 0 L ___
0
J... _ _ _
20
. . .~.
40
_ _- ' - - - . . _ . - - L . . _ - - - - . . . J
60 80 100
roughly 20 for clean sands and 15 for silty Relative Density , Dr (%)

sands, there is a strong need to obtain addi- Fig. 16. Relationship between dynamic
tional evidence to clarify that point. An stress ratio and relative density in
.alternative way to express resistance of soil terms of double amplitude of axial
to liquefaction is a laboratory test on un- strain at the end of 15 cycles in
disturbed samples to determine the soil cyclic triaxial test
resistance in terms of stress ratio and strain
Single Amplitude Sheer Strain, r (%)
level, and its relation to SPT N-values. 2 5 10 20

Because of great difficulty in obtaining high

·quality undisturbed samples of dense sand,
100
the method based on laboratory testing proce-
dure on so-called "undisturbed" samples tends
to underestimate the resistance to liquefac-
tion (Tokimatsu and Yoshimi, 1981), and
the degree of underestimation could increase
0 Triaxial Test c.~ Reconstituted Sample
as the sand becomes denser. .6. Simp Is Sheer Tast on Reconstituted Sample
(DeAlba e: al, 1976 I
A large amount of undisturbed samples of
10 20
dense sand was obtained recently in Niigata Double Amplitude Axial Strain, DA (%)
City by means of an in situ' freezing method Fig. 17. Empirical relationship between
(Yoshimi et al, 1977), which is considered strain and C
most suitable for obtaining high quality
undisturbed samples of clean sand and hence C is an empirical parameter depending on
can yield a most appropriate value of soil strain amplitude. Based on the relationship
strength. Extensive cyclic triaxial tests were for reconstituted sands proposed by Kokusho
conducted on the samples as well as so-called et al (1981), the term C is added in Eq. (11)
"undisturbed" samples obtained by a conven- in order to take in to account the effect of
tional tube sampling procedure. Detailed strain.
procedures and results of the in situ test Because the relationship between C and
.and the laboratory test will be presented strain for reconstituted samples is reasonably
.elsewhere by the authors. linear on a semi-log chart as shown in Fig. 17
Though data are quite limited, Fig. 16 with a=O. 39 and n=l4, the following rela-
shows a relationship between shear stress tionship may be given for in situ sands on
ratio and relative density for these soils in the basis of the laboratory tests as shown m
terms of axial strain amplitude at the end Fig.16, provided a=O. 45 and n=l4.
of 15 cycles. The relationship may be well Ca=97-19log DA
represented by the following equation: for triaxial test (12)
Cs=94-19log r
2::,-=a{l~~+( ~r )n} (11) for simple shear test (13)
in which a and n are empirical constants, and m which DA is double amplitude of axial

NII-Electronic Library Service

 EMPIRICAL CORRELATION OF LIQUEFACTION 67

depending on the methods to determine the

-.- Averaoe Curves for Submerged
Sand without Fine&
maximum and minimum densities. Neverthe-
60 I .
z less, Eq. (14) seems applicable to in situ
~~t~~a~;; ~:~~ w~~vt: o! ~ines
1 0
- /
.. 60 ---Relationship from Eq. (16) ;' clean sands which have probably been sub-
~
0
/ jected to some aging effects. Note that using
~ 40
z SPT N-values by the cathead and rope
f-
~ 20 method with significant rope friction may
overstimate the relative density by 10 to 20
0
0 20 40 percentage points for the sands tested. Sub-
Relative Density, Dr (%)
stituting Eq. (10) into Eq. (14) leads to
Fig. 18. Relationship among SPT N-
values, relative density and ef- Dr=16-/Jf; (15)
fective vertical stress A drastic decrease in SPT N-values for
strain, and r is single amplitude of shear saturated sands with fines compared with
strain. Fig. 16 shows that Eq. (11) together submerged sands without fines, such as that
with Eq. (12) could simulate reasonably well observed in the field case histories, is also
the relationship between stress ratio and apparent in Fig. 18, although it may be over-
relative density of the undisturbed clean estimated because of inevitable boundary con-
sands. ditions in the laboratory test (Gibbs and
On the other hand, relationships between Holtz, 1957). The relationship for this fine
relative density and N-values for sands which sand may be represented by the following
have been studied by several researchers equations as shown in Fig. 18 assuming that
(Gibbs and Holtz, 1957, Marcuson and Bie- L1N1 =15, although the general applicability
ganousky, 1976) are summarized in Fig. 18. of this equation for estimating the relative
For clean sands without fines, the following density of fine sands is yet to be proven.
relationship was given by Meyerhof (1957) :

(14)
Dr= 21;-:::;~~~~~~~~ (16)
or
in which 0 0 1 denotes the effective overburden (17)
stress in kgfjcm 2 (98 kPa). Even though
in which LJN1 is a constant and can be
Eq. (14) is based simply on freshly deposited
considered as a correction term for taking
very young sands in the laboratory tests,
into account the effect of fines content, since
the in situ SPT measurement and laboratory
Eq. (17) is reduced to Eq. (15) if LJN1 =0.
determination of relative density on the
Assuming that the relation between shear
undisturbed sands by the in situ freezing
stress ratio and relative density is uniquely
method cited previously show that the
defined by Eq. (11) irrespective of the
relative density of the sands at o 0 '=1 kgfjcm 2
presence of fines, the following relation
(98 kPa) with SPT N-value of 30 is 80 to 90
can be obtained by substituting Eq. (17)
percent, which is closely related to the value
into Eq. (11).
determined by direct substitution of N and
<1 0 1 into Eq. (14), provided that the relative
density is determined on the basis of the
_!!_!!:__=a[ 16JN;+L1}J7 +(
2 oc' 100
16-JN+1
Ca
LJN1 )n]_
maximum and minimum densities obtained (18)
by the JSSMFE method (JSSMFE, 1979). Eq. (18) can represent directly the relation-
That Eq. (14) could successfully relate the ship between shear stress and SPT N-values
SPT N-value and relative density for the in terms of shear strain.
high quality undisturbed sample may just be The following items must be considered
fortuitous when one remembers the fact that when one attempts to relate the undrained
relative density values may vary considerably strength of in situ saturated sand during an

NII-Electronic Library Service

 68 TOKIMATSU AND YOSHIMI

earthquake and that obtained by the cyclic 0.7 r----r---.---..,.---.----~-,

• Extensive Liquefaction
triaxial test even if the soil sample is per- • Moderate Liquefaction
A Marginal Condition
0. 6 o No Liquefaction
fectly undisturbed: (1) the effect of system r=-2%

compliance including membrane penetration DA" 5% <Y~ 3 8 %)

\2 0.5 ® Lob.TestonUndisturbed
Sample byln Situ freezing
and of end friction, (2) the difference be- }J
tween the triaxial condition and the simple 0.4 0

shear condition, (3) the effect of irregular

shear stress, and ( 4) the effect of multi-
.'2
&0.3
•
.. -Iwasaki et ol
directional shear. However, for lack of direct "'
"' 0.2 --0
-- --(1978b)

~ ~~~ QA:S%0'"3.8%)
en
evaluation of the above using high quality ®Lab J~!~s~u r;:~c;~~:le
0

by Conventional
g Method
undisturbed samples of dense sand, the V5 0.1
correction factor Cr=O. 57 proposed by De Fines Content ~ 5%

Alba et al. (1976) based on their large-scale

Normalized SPT N-value, N1
simple shear tests on a reconstituted clean
sand will be tentatively adopted here, i.e., Fig. 19. Comparison between field cor-
relation and laboratory test results
(OoTz)
1 -c( 2oa)
field- oc'
r triaxial
09 ) for clean sands

0.7
Thus, the relationship between dynamic • Extensive Liquefaction
• Moderate Liquefaction
t:. Marginal Condition
shear stress ratio and SPT Ncvalue with O.G o No Liquefaction

respect to shear strain amplitude for rep- 10% 5% r~2%

resentative samples of in situ soils may be b 0.5

.....
defined by the following equation. ~ 0

0.4

~=aC [16vNl+L1N1 +(16,JN-;+;;TFT;)n] .'2

oo' r 100 Cs ~ 0.3

(20)
where a=O. 45, Cr=O. 57, n=14, L1N1 =0 for
clean sands and L1N1 =5 for silty sands are
assumed. The last assumption is rather
arbitrary but judged by the authors to be
appropriate.
Comparison Between Field Evidence and
Laboratory Test Results
For comparison purposes, the field data
shown previously are replotted in Figs. 19
a 1d 20 together with the curves defined by
Eq. (20). Also shown in Fig. 19 are the
direct relationships between dynamic shear
stress ratio and normalized SPT Ncvalues
for the undisturbed samples obtained in
Niigata City. It can be noted that the data
by the in situ freezing method seem to be
consistent with the field behavior. It may be
apparent from Fig. 19 that the curves based
on the laboratory tests on the samples by the
in situ freezing method could represent the
boundary between liquefaction and no-lique-
faction conditions. It seems reasonable,
Fines Content 1£ I 0%
Normalized SPT N -value, N1
Fig. 20. Comparison between field correla-
tion and laboratory test results for
sands containing more than 10% fines
therefore, to consider that the absence of
extensive damage in sands with SPT Nc
values greater than certain limits, say 20
for clean sands and 15 for silty sands, 1s
due mainly to the inherent strength of
dense sands. This is consistent with the con-
cept of cyclic mobility by Castro (1975) and
that of limiting shear strains by De Alba,
Seed and Chan (1976).
It is noteworthy in Fig. 19 that conven-
tional tube sampling for dense sands appears
to produce significantly lower strength than
those estimated from field behavior, and that
the test result is in fairly good accord with
the criteria proposed by Iwasaki et al.

NII-Electronic Library Service

 EMPIRICAL CORRELATION OF LIQUEFACTION 69

(1978 b) which are based on their laboratory
tests on tube samples. The strength ratios
at DA=5% between the in situ freezing
method and the conventional tube sampling
from the figure would be roughly 1. 0, 1. 2,
1. 5 to 2, and 2. 5 to 3 for SPT N 1-values of
15, 20, 25 and 30, respectively, which indicate
that considerable underestimation of un-
drained strength does occur for medium
dense to dense sands if it is based on
laboratory tests on samples by conventional
methods.
Despite many assumptions for fine sands,
it may also be noted from Fig. 20 that the
curves in the figure happen to represent the
boundary between liquefiable and non-lique-
fiable conditions, although the shear strain
amplitude may not correspond to the actual
one.
Eq. (20) and Figs. 19 and 20 indicate that
the effect of presence of fines defined by Eq.
(20) is only shifting the boundary curve to
the left by a distance of iJN1 from the curve
Table 5. Correlation factors
for determining SPT
Na-value
Fines Content
FC (%)
-----------------
0 - 5 with more than 60 percent fines and gravel
5 - 10
10 -
0.7 ,.--..--..---r---r--..,....--,
• Extensive Liquefaction
• Moderate Liquefaction
with ilN1=0 and that the curve defined by
ilN1 =5 agrees remarkably well with a dif-
ference of mean values between clean sands
and sands containing some fines as shown in
Table 4. In order to eliminate ilN1 in Eq.
(20), an adjusted SPT Na-value is defined by
Na=N1+ilN1 (21)
in which iJN1 is a correction term defined
in Table 5, which is assumed from Figs. 19
and 20, and Table 4. Substituting Eq. (21)
into Eq. (20) leads to
~~ =aC{-!.6(o~~=+( 16~~:-)n] (22)
Fig. 21 shows the final results for field obser-
vations using the adjusted SPT Na-value
with boundaries from Eq. (22). Note that in
using this value, liquefaction condition ap-
pears to be uniquely determined regardless of
the fines content. Also noted are the facts
that the boundary between liquefaction and
no-liquefaction conditions corresponds to the
line determined by the shear strain ampli-
tude of two percent, and that the boundary
for extensive liquefaction seems to correspond
to the line defined by a 10 percent shear
strain.
~Nf
Because the data scarcely include silty sands
Interpolate soils with more than 30 percent gravel,
0.1FC+4 application of the method to these soils is
restricted, but it would provide a conserva-
tive estimation for silty sands as far as using

0.6
tJ. Marginal Condition
o No Liquefac1ion
the values listed in Table 5.

10%6"/o )'::2"'/.,
~ 0.5
0

tJ' FACTOR OF SAFETY AGAINST SEIS-

0.4 MIC DAMAGE INVOLVING LIQUEF AC-
.9 TION
0
a:: 0.3
Factor of safety against Liquefaction
&l
~
0
Factor of safety against liquefaction, Fl, is
iii 0.2
6
usually defined in terms of shear stress ratio
Cl)
.<::
(J) 0.1 by the following equation:
0 o~~I0~~2~0-~3~0-~4~0--~5~0~ F l_ (Tz/cfo') at a given shear strain r
- ------- -------------------·- - - - - - - - - (23)
Adjusted SPT N-volue, No
(Td/cfo')
Fig. 21. Relationship between shear stress Note that what Fl=1 means depends on the
ratio and adjusted SPT Na-value for SPT N 1-values which may be apparent in
various soils with gravel content less Fig. 19, i.e., if the SPT Ncvalue is only
than 25% 10 and Fl=l. 0 at a given shear strain of two

NII-Electronic Library Service

 70 TOKIMATSU AND YOSHIMI
percent, a 10 percent increase in the denom-
inator of Eq. (23) may cause far more than
10 percent shear strain which leads to a
catastrophic failure. On the contrary, even
though the same increase in the shear stress
occurs in sands with the SPT N 1-value of
30, the induced shear strain will still remain
around two percent which is far from cata-
strophic failure. Since earthquake ground
motions cannot be estimated within an
accuracy of 10 percent, it is preferable that
a fairly large factor of safety such as 1.5 or
more should be used for low Ncvalues, e. g.,
N 1 ;;£ 10. On the other hand, one might
adopt a slightly lower factor of safety such
as 1. 3 or less for medium dense to dense
sands.
Pore Pressure Buildup in Sand Where Lique-
faction Does not Occur
Even if the shear strain in sand dose not
exceed a critical value or the pore pressure
does not increase to a value equal to the
initial effective stress, i.e., the safety factor,
Fl, is greater than unity, it may be desirable
to estimate the degree of pore pressure
buildup due to a given earthquake. A sim-
plified procedure to estimate the pore pressure
generation is presented herein on the basis
of empirical equations concerning liquefac-
tion.
Seed et al. (1976) showed that the pore
pressure buildup in saturated sand under
undrained conditions could be expressed by
ru=O. 5+sin- 1 (2 rNI/a_l)/ll (24)
m which ru=ufao' (u=excess pore water
pressure), rN=the ratio of the number of
cycles, Ni, to Nl that corresponds to ru=1,
and a is an empirical constant. The effect
of the magnitude of a on the relationship
between ru and rN is defined elsewhere
(Tokimatsu and Yoshimi, 1982). Since a
relationship between shear stress amplitude
and number of cycles which determines
liquefaction resistance of loose saturated
sands under undrained conditions 1s ap-
proximately straight on a log-log plot,
the following equation can be derived.
Foetor of Safety for Liquefaction, F,
Fig. 22. Relationship between pore pres-
sure ratio and factor of safety
against liquefaction
rN=NtfNz= (Tt/Tz) 1113 =Fz 11 f3 (25)
m which {3 is the inclination of the line on
a log-log plot. Substitution of Eq. (25) into
Eq. (24) leads to
rt~=O. 5+sin- 1 (2Fzllaf3_1)/7t (26)
Note that this equation is valid only if Fl;;;;).
Because representative values of a and {3
are around 1 and -0. 2, respectively, a
typical relationship between ru and Fz can
be obtained by substituting 1fa{3= -5 into
Eq. (26), which is shown in Fig. 22 together
with the other probable values. A similar
estimation but based on the relationship
between ru and safety factor in terms of
number of cycles has recently been proposed
by Seed and Idriss (1981).
Effects of SPT Method on Estimation of
Liquefaction Resistance
Because it is conceivable that the criteria
presented herein is based on SPT N-values
determined by the free fall method in which
the ratio of net impact energy dilivered to
the rod is probably close to 0. 8, appropriate
correction should be made when using these
criteria with SPT N-values determined by
other methods in which the energy ratio
may be different.
For comparison purposes, Fig. 23 shows the
boundary lines separating liquefaction and no
liquefaction conditions based on N 1 values by
the free fall type method in Japan, Nti• the
cathead and rope methods in Japan Nci• and
the cathead and rope method in the U.S. A.,
Ncf· It can be seen in the figure that the
NII-Electronic Library Service

0.7 .---~----,-----,.----r-----r---,
I
0.6
i2 0.5 <:l:i<!IJ"f(]~ti_()~
h;'
.i 0.41
~ 0.3
~
Vj 0.1
0
0 10 20 30
Normalized SPT N-value, N 1
Fig. 23. Boundary curves separating lique-
faction and no liquefaction conditions
based on three varieties of the SPT :
the free fall type method in Japan,
NtJ, the cathead and rope method in
Japan, NcJ• and the cathead and rope
method in the U.S. A., Ncf
difference in the resistance for a g1ven N-
value is significant and cannot be neglected
particularly when the SPT N-value exceeds
about 20, which undoubtedly indicates the
need of energy correction when one uses
the cathead and rope method without any
efforts to reduce rope friction ; otherwise
the estimation of soil resistance from the
relationship such as Fig. 21 could be on the
unsafe side.
CONCLUSIONS
On the basis of a review of field behavior
during several earthquakes together with
laboratory tests on high quality undisturbed
samples of sand, the following conclusions
may be drawn:
( 1) Sands containing more than 10 per-
cent fines have much greater liquefaction
resistance than clean sands having the same
SPT N-values.
( 2) Extensive damage due to liquefaction
would not occur for clean sands with SPT
N 1-values greater than 25, and silty sands
containing more than 10 percent fines (finer
than 74.um) whose SPT N 1-values are greater
than about 20.
EMPIRICAL CORRELATION OF LIQUEFACTION 71
( 3 ) Soils containing more than 20 per-
cent clay (finer than 5 ,urn) would hardly
liquefy unless their plasticity indexes are low.
( 4) Sands containing gravel particles
seem to have less resistance to liquefaction
than clean sands without gravel having the
same SPT N 1-values.
( 5) Since the SPT N-value determined
by the cathead and rope method in Japan is
greater by about 20 percent than that mea-
sured by the trip monkey method, the me-
5% Shear Strain thod of SPT procedure should be taken into
40 50 account in evaluating liquefaction resistance
particularly when SPT N 1-values exceed 20.
Finally the following procedure for esti-
mating soil liquefaction potential of level
ground may be proposed based on the above
conclusions :
( 1) If clay content is greater than 20
percent consider that the soil does not liquefy
unless its plasticity index is low or the soil
has high sensitivity.
( 2) Assume equivalent dynamic shear
stresses within a given site by
~=0.1(M-1) amax ~(l-0.015z)
(fo' g (fo'
(6)
( 3) Correct N-values in terms of net
energy delivered to the rod by Eqs. (8) and
(9), if necessary.
( 4) With Table 5, compute adjusted Na-
values by
(21')
( 5) Compute soil liquefaction resistance
by Fig. 21 or by
::, =ac{lf!_(o[;a -+(_1 6 ~~~)n] (22)
in which a=O. 45, Cr=O. 57, n=14, and
Cs=80""'90 for separating liquefaction and no
liquefaction conditions and C 8 =75 for exten-
sive liquefaction.
( 6) Compute factor of safety against
liquefaction by
Fz= (-rzl(fo')/(-rdf(fo') (23')
Note that this method should not be used
for soils with static shear stresses on hori-
zontal plane because almost all the data in
NII-Electronic Library Service
 72 TOKIMATSU AND YOSHIMI
this paper are based on field performance of
relatively level ground with free surface.
One must be particularly careful with loose
soils with static shear stresses on horizontal
plane because their liquefaction resistance is
likely to be overestimated by direct appli-
cation of this method.
ACKNOWLEDGEMENTS
This study was partially supported by a
research grant from the Ministry of Educa-
tion, Science and Cui ture. Valuable discus-
sions by Professor H.B. Seed of the University
of California, Berkeley, during the prepara-
tion of this paper are greatly appreciated.
State-of-the-art in Mainland China concern-
ing the SPT method and estimation of maxi-
mum horizontal accelerations provided by
Mr. Z. Zhang, Visiting Scholar, University of
California, Berkeley, is gratefully acknowl-
edged.
REFERENCES
1) Bennett, M. J., et al (1981) : "Subsurface in-
vestigation of liquefaction, Imperial Valley
Earthquake, California, October 15, 1979,"
U. S. Geological Survey Open-File Report
81-502, 83 pp.
2) Castro, G. (1975) : "Liquefaction and cyclic
mobility of saturated sands," Jour. Geotech.
Eng. Di v., ASCE, Vol. 101, No. G T 6, pp.
551-569.
3) De Alba, P., Seed, H. B. and Chan, C. K.
(1976) : "Sand liquefaction in large-scale
simple shear tests," Jour. Geotech. Eng.
Div., ASCE, Vol. 102, No. GT9, pp. 909-927.
4) Dixson, S. J. and Burke, J. W. (1973) : "Lique-
faction case history," Jour., Soil Mech. and
Found. Div., ASCE, Vol. 99, No. SM 11,
pp. 921-955.
5) Finn, W. D. L. (1982) : "Soil liquefaction stud-
ies in the People's Republic of China," Soil
Mechanics-Transient and Cyclic Loads, Ch.
22, pp. 609-626, John Wiley & Sons, Ltd.
6) Gibbs, H. J. and Holtz, W. G. (1957) : "Re-
search on determining the density of sands
by spoon penetration testing," Proc., 4th Int.
Con£. Soil Mech. and Found. Eng., London,
Vol. 1, pp. 35-39.
7) Iida, K. (1977): "Distribution of damage and
earthquake intensity during the Tohnankai
Earthquake of 1944," Committee for Disaster
Prevention in Nagoya Prefecture, 1977-3 (in
Japanese).
8) !ida, K. (1979) : "Distribution of damage and
earthquake intensity during the Mino-Owari
Earthquake of 1891," Committee for Disaster
Prevension in Nagoya Prefecture, 1979-3 (in
Japanese).
9) Ishihara, K., Silver, M. L. and Kitagawa, H.
(1979) : "Cyclic strength of undisturbed sands
obtained by a piston sampler," Soils and Foun-
dations, Vol. 19, No.3, pp. 61-76.
10) Ishihara, K., Kawase, Y. and Nakajima, M.
(1980) : "Liquefaction characteristics of sand
deposits at an oil tank site during the 1978
Miyagiken-Oki Earthquake," Soils and Foun-
dations, Vol. 20, No.2, pp. 97-111.
11) Ishihara, K., Shimizu, K. and Yamada, Y.
(1981) : "Pore water pressures measured in
sand deposit during an earthquake," Soils and
Foundations, Vol. 21, No.4, pp. 85-100.
12) Iwasaki, T., Kawashima, K. and Tokida, K.
(1978 a) : "Report of the Miyagiken-Oki
Earthquake of June, 1978," Public Works
Research lnsti tu te, Ministry of Construction,
Report No. 1422, 202 pp (in Japanese).
13) Iwasaki, T., et al. (1978 b) : "A practical
method for assessing soil liquefaction potential
based on case studies at various sites in
Japan," Proc. 2nd Int. Con£. on Microzo-
nation, San Francisco, Vol. 2, pp. 885-896.
14) Iwasaki, T., et al. (1981) : "Studies on aseis-
mic stability of river dykes considering ground
liquefaction," Proc., 16th Annual Meeting,
JSSMFE, pp. 637-640 (in Japanese).
15) Kishida, H. (1966) : "Damage to reinforced
concrete buildings in Niigata City with special
reference to foundation engineering," Soil and
Foundation, Vol. 6, No.1, pp. 71-88.
16) Kishida, H. (1969) : "Characteristics of lique-
fied sands during Mino-Owari, Tohnankai and
Fukui Earthquakes, " Soils and Foundations,
Vol. 9, No.1, pp. 75-92.
17) Kishida, H. (1970) : "Characteristics of lique-
faction of level sandy ground during the
Tokachioki Earthquake," Soils and Founda-
tions, Vol.10, No.2, pp.l03-111.
18) Kodera, J. (1964) : "Earthquake damage and
the ground of pier foundations, Part 1,"
Tsuchi-to-Kiso, Vol.12, No.3, pp.11-18 (in
Japanese).
19) Koizumi, Y. (1966) : "Changes in density of
NII-Electronic Library Service
 EMPIRICAL CORRELATION OF LIQUEFACTION
sand subsoil caused by the Niigata Earth-
quake," Soil and Foundation, Vol. 6, No.2,
pp. 38-44.
20) Kokusho, T., Kato, S. and Shimada, M.
(1981) : "The effect of initial shear stress on
cyclic shear characteristics of saturated dense
sand," Proc., 16th Annual Meeting, JSSMFE,
pp. 581-584 (in Japanese).
21) Kovacs, W. D., Evans,]. C. and Griffith, A. H.
(1977) : "Towards a more standard SPT,"
Proc. 9th Int. Con£. Soil Mech. and Found.
Eng., Vol. 2, pp. 269-276.
22) Kovacs, W. D. and Salomone, L. A. (1982) :
"SPT hammer energy measurements," Jour.
Geotech. Eng. Div., ASCE, Vol.108, No.
G T 4, pp. 599-620.
23) Lee, K. L. and Fitton, J. A. (1968) : "Factors
affecting the cyclic loading strength of soil, "
Vibration Effects of Earthquakes on Soils and
Foundations, ASTM STP 450, Am. Soc. for
Testing and Materials, pp. 71-95.
24) Lee, K. L., et al. (1975) : "Properties of soil
in the San Fernando hydraulic fill dams,"
Jour. Geotech. Eng. Div., ASCE, Vol. 101,
No. GT 8, pp. 801-821.
25) Marcuson, W. F. III, and Bieganousky, W. A.
(1976) : "Laboratory standard penetration tests
on fine sands," Jour. Geotech. Eng. Div., Vol.
103, No. GT 6, pp. 565-588.
26) Meyerhof, G. G. (1957) : Discussion, Proc.
4th Int. Con£. on Soil Mech. and Found.
Eng., Vol. 3, p. 110.
27) Nishizawa, Y., Fuyuki, S. and Uto, K. (1980):
"In situ tests and efficiency concerning SPT,"
Proc., Sounding Symp., JSSMFE, pp. 83-92
(in Japanese).
28) Ohsaki, Y. (1966) : "Niigata Earthquake,
1964, building damage and soil conditions,"
Soil and Foundation, Vol. 6, No.2, pp. 14-37.
29) Ohsaki, Y. (1970) : "Effects of sand compact-
ion on liquefaction during Tokachioki Earth-
quake, " Soils and Foundations, Vol. 10, No.
2, pp. 112-128.
30) Okusa, S., Arima, S. and Maikuma, H. (1980):
"Liquefaction of mine tailings m the 1978
Izu-Ohshima-Kinkai Earthquake, Central Ja·
pan," Proc. 7th World Con£. on Earthquake
Eng., Istanbul, Vol. 3, pp. 89-96.
31) Pyke, R. M., Knuppel, L.A. and Lee, K. L.
(1978) : "Liquefaction potential of hydraulic
fills," Jour. Geotech. Eng. Div., ASCE, Vol.
104, No. GT 11, pp. 1335-1354.
32) Ross, G. A. (1968) : "Case studies of soil
stability problems resulting from earth-
73
quakes," Thesis submitted to the Univ. of
Cal., Berkeley.
33) Schmertmann, J. H. (1978) : "Use the SPT to
measure dynamic soil properties? -yes, but
· · ·!," Dynamic Geotech. Testing, Am. Soc.
for Testing and Materials, STP 654, pp.
341-355.
34) Seed, H. B. and Idriss, I. M. (1971) : "Simpli-
fied procedure for evaluating soil liquefaction
potential," Jour., Soil Mech. and Found. Div.,
ASCE, Vol. 97, No. SM 9, pp.1249-1273.
35) Seed, H. B., et al. (1975) : "Representation of
irregular stress time histories by equivalent
uniform stress series in liquefaction analyses, "
Earthquake Eng. Research Center, Univ. of
Cal., Berkeley, No. 75-29.
36) Seed, H. B., Martin, P. P. and Lysmer, ].
(1976) : "Pore-water pressure changes during
soil liquefaction," Jour. Geotech. Eng. Div.
ASCE, Vol.102, No. GT 4, pp. 323-346.
37) Seed, H. B. (1976) : "Evaluation of soil lique·
faction effects on level ground during earth-
quakes," Liquefaction Problems m Geotech.
Eng., ASCE Annual Convention and Exposi-
tion, held at Philadelphia, pp. 1-104.
38) Seed, H. B. (1979) : "Liquefaction and cyclic
mobility evaluation for level ground during
earthquakes," Jour. Geotech. Eng. Div.,
ASCE, Vol.105, No. GT 2, pp. 201-255.
39) Seed, H. B. and Idriss, I. M. (1981) : "Evalua-
tion of liquefaction potential of sand deposits
based on observations of performance in pre-
vious earthquakes," Preprint 81-544, In Situ
Testing to Evaluate Liquefaction Suscepti-
bility, ASCE National Convention, St. Louis,
Missouri, October 27.
40) Seed, H. B., et al. (1981) : "Earthquake m-
duced liquefaction near Lake Amatitlan,
Guatemala," Jour. Geotech. Eng. Div.,
ASCE, Vol.107, No. GT4, pp. 501-518.
41) Takada, T., et al. (1965) : "A report of Niigata
Earthquake-Part 5, damage of bridges,"
Report No. 125-5, Public Works Research
Institute, Ministry of Construction.
42) Tokimatsu, K. and Yoshimi, Y. (1981) :
"Field correlation of soil liquefaction with
SPT and grain size," Proc., Int. Con£. on
Recent Advances m Geotech. Earthquake
Eng. and Soil Dynamics, St. Louis, Mo.,
Vol. 1, pp. 203-208.
43) Tokimatsu, K. and Yoshimi, Y. (1982) :
"Liquefaction of sand due to multidirectional
cyclic shear," Soils and Foundations, Vol. 22,
No.3, pp. 126-130.
NII-Electronic Library Service
 74 TOKIMATSU AND YOSHIMI
44) Tsuchida, H., et al. (1979) : "The damage to
port structures by the 1978 Miyagiken-Oki
Earthquake, " Technical Note, The Port and
Harbour Research Inst., Ministry of Trans-
port, No. 325 (in Japanese).
45) Yoshimi, Y., Hatanaka, M. and Oh-oka, H.
(1977) : "A simple method for undisturbed
sampling by freezing," Proc., Specialty Session
2 on Soil Sampling, 9th Int. Con£. on Soil
Mech. and Found. Eng., Tokyo, pp. 23-28.
46) Yoshimi, Y., Kaneko, 0. and Tokimatsu, K.
(1983) : "Cyclic shear characteristics of a dense
sand sampled by in situ freezing," Proc.
18th Annual Meeting, JSSMFE, pp. 559-560
(in Japanese).
47) Yoshimi, Y. and Tokimatsu, K. (1983) :
"SPT practice survey and comparative tests,"
Soils and Foundations, Vol. 23, No.3, pp.
105-111.
48) Zhou, S. G. (1981) : "Influence of fines on
evaluating liquefaction of sand by CPT,"
Proc., Int. Con£. on Recent Advances in Geo-
tech. Earthquake Eng. and Soil Dynamics,
St. Louis, Mo., Vol.1, pp.167-172.
NII-Electronic Library Service
Another random document with
no related content on Scribd:
speak to Your Honour,’ replied he, ‘I own I am taken in an unhappy
affair, which neither malice nor inclination drew me into. But the
Head of the Country, notwithstanding I was the only child of my aged
parents, insisted, under great penalties, on my coming in to join him;
so that I beg you will pity me in my condition: but if it must be that I
die this night, pray! tell me what death you think it will be?’ Upon
which, not knowing what to think or what to say, I was somewhat at a
loss: his telling me of his aged parents, and his simplicity touched
me much: and how far I may be censured for my after-conduct, I
know not; but those who think I did amiss, I hope, will pardon me, as
being then but a young warrior. True, such a thing, I know, ought not
to have been publicly done; yet, when an opportunity presented itself
of doing an Act of Mercy without harm to the Cause, I am convinced
that this my behaviour will, with all generous minds, escape reproof.
The young boy continuing his lamentations, I told him to be of good
courage, for death was not so imminent as he expected; at which he
seemed to respire new vigor and life: And after some questions
asked, and promises made of his ever being grateful and upon his
telling me, that his home was not above fourteen miles from thence, I
asked him, if he could privately slip away thither. He replied, such a
favor would be too great for him to presume to request: So telling
him to do it if he could, I discharged him, ordering him to be a
support to his poor aged parents.
And now by this time we had almost conducted our prisoners to
Falkirk; and, after delivering them, we went thoroughly wet and cold,
to repose ourselves a little while on straw, and some in the open
fields or air, all places being by this time entirely filled, so that it was
then impossible to find any resource or ease for our excessive
hunger, wet and cold. So resting a few hours in that condition, they
soon after appeared; and when light, we went to see the field of the
slain; and number being made of them, we found near seven
hundred of the enemy slain, and about fifty of ours, which were
immediately interred. But this cheap-bought victory, you will say,
merited a better exit! The most distinguished among the slain were
Colonel Whitney and Sir Robert Monro, who was heard much to
blaspheme during the engagement, and as a punishment for which,
his tongue was miraculously cut asunder by a sword, that struck him
directly across the mouth. His brother, a physician, was likewise
killed at his side.[423] There were likewise found slain, some
Presbyterian Parsons,[424] who, fired with holy zeal, had quitted their
Bibles and took their swords. It was said, that one of these Parsons,
seeing the danger he was in of losing his life as a Soldier, had
recourse to his dignity, supposing that would be a cloak to save him.
‘Spare my life,’ said he to a Highlander, who was on the point of
taking it, ‘for I am a Minister of My Master Jesus Christ!’ To which the
other ingeniously replied: ‘If you are a good one, your Master has
need of you; if not, it’s fitting that you go and take your punishment
elsewhere!’—which dilemma was immediately solved by the
Highlander’s sword. Another Minister, seeing the case his Brother
was in, and being in a fair way to share the same fate, begged his
life of another Highlander for Prince Charles’s sake, by which means
he preserved what he would otherwise inevitably have lost.
We now took possession of the enemy’s baggage, camp, and
eight pieces of cannon; which they had not time to carry off—besides
a few prisoners, the chief of whom was Major Lockhart, who, after
having his life given him, and his liberty upon his parole of honour,
afterwards spurned against gratitude itself, by not only being
heinously perjured, but more than ordinarily thirsty of those
prisoners’ blood who just before had spared his life.[425]
We had now about one-hundred prisoners, one of whom seeing
his situation and ours, said with a grave countenance to his
companion: ‘By my soul, Dick, if Prince Charles goes on in this way,
Prince Frederick will never be King George!’ But of the five or six
thousand men that went with the Prince to the field of battle, scarce
three thousand returned back with him, for many of them, having
loaded themselves with booty, returned up to the hills. It was now a
great loss to us, that we had neither fort or other secure place to
keep our prisoners in, so that, if it were not merely out of mercy, it
was to no purpose to take prisoners, being without the means of
keeping them.
And now being come again to Stirling, the enemy being fled to
Edinburgh, and finding the siege of the Castle went on but slowly the
spirit of the army began much to change. Factions, grudges, and
private interest were now judged proper to be exercised, so that the
Prince was in a mortifying situation, seeing himself deserted by half
his army, and the others mightily turned. But as it is not for me to say
more than what relates to myself, I shall only mention the grudge I
have often dearly paid for—that exercised against my Colonel, who,
on the death of his brother had lately succeeded to the title of Lord
Balmerino. And here it may be proper to give some account of the
character of this brave man, which though sufficiently known, his
praises cannot enough be sung. He was a man of a noble
personage, of approved loyalty, and had the courage of a lion. He
possessed a mind and genius well ornamented with both natural and
acquired parts, being versed in most languages. He could recite
whole pages of Horace, Ovid, and Virgil at his age of fifty-eight, as
perfectly as a school-boy of sixteen, so that his memory for his years
was wonderful, the more so for its not being in the least impaired by
his hard drinking—his sole and predominant passion, which if he had
a little more restrained, he would have shone with the same lustre in
the army as he afterwards did on the scaffold. However, whilst with
the Prince, he was never failing with his duty; and proud he was of
having something, wherein he could shew his loyalty and obedience
to his Master.
But what was the cause of the quarrel betwixt him and Lord
George Murray, I know not; as Lord Balmerino did not ever directly
inform us why he was treated after that manner: only I remember,
His Lordship, when he saw himself so apparently ill-used, frequently
addressed his corps as follows. ‘Come let us do as we are ordered!
It is in vain to dispute; a time will come when I shall see all these
things righted, and that too at Lord George’s cost or mine. But at
present he is my superior, and we must obey as we tender the good
of the Prince.’ With such soothing expressions has he often accosted
us, when some were mutinying. And certainly he suffered a great
deal from Lord George, for, to my knowledge, we have been ordered
twenty or thirty miles, harassed and fatigued, and a courier sent after
us, ordering us immediately to return without halting—saying it was
my Lord George Murray’s orders.[426] And now a harder time than
ever came upon us, for I can safely say and prove, that from the time
of Falkirk Affair to the Battle of Culloden, notwithstanding the fatigue
of the day, I scarce slept three nights out of seven in bed.
And now having been some time before Stirling Castle, news
arrived from Edinburgh, that Cumberland was come thither with an
intention to rally General Hawley’s army and attack us again. As our
number had been so much diminished by the desertion of those who
after the battle of Falkirk had gone home loaded with plunder; it was
judged expedient for us to retire higher into the country, where we
were sure of being joined by some more forces. Accordingly we had
orders to proceed on our march; and on the morning on which we
began it, the Prince to St. Ringin,[427] distant about half a mile from
Stirling, to give the necessary directions for quitting the town and
raising the siege; which being done, we retired again, and when at a
short distance, we were surprized with a hideous noise just behind
us; and upon the Prince’s sending back to inquire what it was, it was
found to be a church blown up, just upon the place where a few
minutes before we had been standing. This church had been
converted into a magazine for our gun-powder, which by some
accident had been set on fire, and several of the town’s-people and
of our soldiers were killed by the explosion. However, continuing our
march, we arrived that night at Crieff a little town in the Highlands;
and the Prince lodged at Drummond Castle, the residence of the
illustrious Duke of Perth, which was only a short distance from the
said town. And now it was judged proper for the army to separate;
accordingly the Prince went the Highland way with the Highlanders
and prisoners for Inverness; and the Horse and Lowland Regiments
the Low-Country Road by the Sea-Coast, which was much longer, so
that it was three or four weeks before we again formed a junction.
And now we marched from Crieff to Perth, a large fine town; from
Perth to Cowpar-in-Angus; from Cowpar to Glams; from Glams to
Forfar; and so on to Montrose. The reason why I am so short in
mentioning these places, is, I have little to say; we passing them
quickly, and nothing extraordinary happening. But at Montrose we
halted a few days. It is a fine loyal seaport town and looked upon as I
was told there, to produce men of the greatest wisdom in Scotland.
Having staid three days there, we were a little surprised at the sight
of ships of war, that appeared a little distance off the Coast: and the
rumor being that Cumberland was pretty nigh us, we began to
apprehend lest they should land and intercept our passage; to
prevent which, we marched with all haste out of the town, the foot
going out the third day at night, and the rest following early the next
morning except the Hussars who remained some days longer to
watch the enemy’s motions.
Some of my readers may be curious to know what sort of men
these Hussars were. A set of braver fellows it would be very hard to
find; many of them having mounted themselves on horses which
they had taken from the enemy. Their Commander[428] also was a
wise, courageous virtuous man, and behaved himself in his station to
the admiration of all, regulating his corps with such order as to make
our enemies and the country, even fifty miles distant from us, have
more fear of them than almost the whole army. In fine, he was of
infinite service to the Prince, as also were his horse; for their conduct
was daring, and their courage was steeled, and few of them there
were, who would have scrupled to go, if possible to hell’s gates to
fetch away the keys.
Soon after our departure from Montrose, we arrived at Aberdeen,
where we staid two or three days; and notwithstanding our being in
the town the Presbyterian Ministers ceased not to preach and pray
publicly against us. And here it was that I happened to be bedfellow
to my Colonel, Lord Balmerino in the same bed where Cumberland
afterwards lay, it being in one of the Chief Provosts’ houses.
When we marched out of Aberdeen, it blew, snowed, hailed, and
froze to such a degree, that few Pictures ever represented Winter,
with all its icicles about it, better than many of us did that day; for
here men were covered with icicles hanging at their eyebrows and
beards; and an entire coldness seizing all their limbs, it may be
wondered at how so many could bear up against the storm, a severe
contrary wind driving snow and little cutting hail bitterly down upon
our faces, in such a manner that it was impossible to see ten yards
before us. And very easy it now was to lose our companions; the
road being bad and leading over large commons, and the paths
being immediately filled up with drifted snow. However, we continued
marching on till about three in the afternoon, when my horse
overcome by the inclemency of the weather, fairly gave it up and
would carry me no farther: and now by a little halt I quickly lost my
Company, and was obliged to alight and lead my horse leg-deep in
snow; being upon a wide common, as it appeared to me, not having
seen all that day’s march scarce a house, tree or barn. I now
expected every moment to perish, as I was quite benumbed and all
covered over with snow, and my horse refused to follow me. At last,
unable to proceed any farther, I sunk down, quite exhausted, upon
the snow. In this dreadful situation, I luckily recollected a little bottle
of spirits, which had been given me by the Provost’s Lady; and
accordingly tried to put my frozen hand into my riding-coat pocket to
take it out. On taking a draught of the liquor, I soon found that never
repose to the wearied traveller, never meat to a most ravenous
hunger, never drink to a most burning thirst, could be more
refreshing or agreeable than this was to me; and I should have
finished my bottle, if a reflection had not come into my head about
my poor horse, which seemed to be in as bad a situation as myself,
being one of a delicate and tender breed. Knowing that he could
drink beer, I resolved to make an experiment whether he would
partake of the contents of my bottle. So pouring the remainder of the
liquor into the crown of my hat, which I had pressed low down for the
purpose, and dissolving some snow in it, in order to mitigate the
spirits, I gave it my poor horse to drink: which to my great surprise
and pleasure, he did, his mouth, I believe, being so cold that he did
not know what he drank. However, now finding ourselves to respire
as it were new life and vigor, we endeavoured to proceed, and after
three quarters of an hour, being almost upon the relapse again, we
stumbled upon a house, and following the walls of it came to the
door, where entering together with my horse, I surprized the poor
people who were sitting at the fire. But they, seeing the condition I
was in, received me with a great deal of good-nature, and permitted
my horse to stand in the house till he was well rubbed, and then led
him to a little place for him to lie in, giving him hay and corn as he
wanted. After I had taken off my riding-coat and boots and well
warmed and refreshed myself, I heard of two more (who had been in
almost as bad a situation as myself except that their horses
continued to carry them), who were come to the next house
adjacent. So going out to see them, I found them to be two old
acquaintances, vizt., one Mr. Maxwell, and Mr. Ball, an English
Gentleman, who, after some chat, concluded, that two more
harassing marches, than that over the Esk and the present, could
scarcely be imagined. Accommodating ourselves as well as we
could for that night, being obliged to lie with our horses, we departed
early next morning for Old Meldrum, which place most of the army
had reached before us.
From Old Meldrum we marched the next day for Banff, a little
pretty agreeable town. About this time we heard, that the
Highlanders, who were with the Prince, had broke down the barracks
of Riven of Badenoch (which were a great eye-sore to them, having
been built to keep them in order), and having taken the Sergeant and
eleven more prisoners, were in full march for Inverness, distant
thence only twenty-four miles. And now, after a short stay at Banff,
we marched for Cullen; and by this time we heard that the Prince
had made himself master of the Lord President’s House, and after
some little resistance forced the soldiers in the town of Inverness to
retire into the Castle, which, after a regular siege, likewise soon
surrendered at discretion. About 200 men were taken prisoners here,
and several officers, the principal of whom were the Governor and
the Master of Ross. This rapid success of the Prince gave us great
courage. So, marching from Cullen, through Fochabers, over the
River Spey, Elgin, Forres and Nairn, towns only ten miles distant
from each other, we came to be greedy spectators of our dear Prince
again, and what he with his brave Highlanders had effected. And
now, to second our victorious arms, we were joined by several more
of the Clans and Chiefs, and the brave Amazon Lady McIntosh,[429]
Seaforth and others coming in, or causing their Clans to come in,
and many who had left us at Falkirk rejoining their colours, greatly
reinforced the army. When in this flourishing condition, it pleased the
Prince to make a visit to the Duke of Gordon’s, whither all the
Guards were ordered to attend him; and in going a curious fine
standard with this motto ‘Britons, strike home!’ that was taken at
Falkirk from Gardiner’s Dragoons, was honourably conferred upon
me by the Prince at the head of the whole Troop; and I had the
honor of carrying it ever after. Having passed Nairn, Forres, Elgin,
and the river Spey, just at the other side under Fochabers, we came
to the Duke’s seat.[430] What reception the Prince had, or what
passed there I know not: but after a short stay we attended our
Royal Master back again to Inverness. But as soon as we arrived
there, we had orders to conduct some of the chief officers, who had
lately been taken prisoners to Forres, and after leaving them there
(they being upon their parole of honor) to go to Cullen to inspect all
affairs there on that side of the country, and to observe the motions
of the enemy who lay at Aberdeen. While at Cullen, where we were
continually upon the watch by our patrolling parties, we heard that
Colonel Grant, a good French Officer, had with his Artillery taken
Fort Augustus, and made the garrison prisoners; that a part of Fitz-
James’ Horse had landed and joined the Prince;[431] and that Lord
George Murray had blocked up seven hundred men in the Castle of
Blair, when he behaved with a great deal of conduct and policy, for,
he appeared before the Castle with only a few men, having hid most
of the Highlanders with him in an adjacent wood. The garrison being
thus deceived made a sally upon him with about three hundred men;
but he immediately drew his party out of the wood and surrounded
them, upon seeing which, they immediately surrendered.[432] So,
sending them prisoners to Inverness he persevered in the siege of
the Castle: but the approach of the Hessians soon obliged him to
raise it and leave four hundred men remaining in the Castle.
At this time President Forbes together with Lord Loudon, were
endeavouring to confederate and knit together what forces they
could: but to prevent their doing much mischief, a party of the brave
McDonalds, with some few others, were detached under the
Command of the Duke of Perth, who soon dispersed this rising
power, and obliged its heads, viz., Lord Loudon and the President to
save themselves in boats.
Some of my readers may be curious to know who this President
was, and what interest he had. He may truly be styled the Oracle of
his Country, for many resorted to him for advice; and had he been as
great a friend as he was an implacable enemy, James would in all
probability have swayed the English sceptre; for by his interest,
cunning and persuasion he brought over his own party, together with
Sir Alexander McDonald and several others, who before were just
sworn in to the Prince’s interest.[433] So, I say, had he been as firm a
friend as he was an implacable enemy, we should have seen,
instead of the four thousand men who marched into England, an
army of Eighteen or twenty thousand men.
About this time we heard, that the officers who were prisoners
upon their parole of honour had broken it, and escaped to the
enemy’s army, all of them except the Master of Ross and one or two
more.[434] But what will the world say, to see these officers, whom no
tie of religion, gratitude or honour could bind, protected and
cherished by their own party—nay sent against us, to endeavour to
destroy those who before had saved them!
And now we heard, that the Advanced Guard of the enemy was
approaching us, and were got up to Strathbogie. We who were also
of the advanced Guard, upon receiving this intelligence, quitted
Cullen, and retreated to Fochabers, where a considerable body of
our men were endeavouring to make a resistance at the River Spey,
and had for that purpose built barracks, and made all necessary
preparations, in case the passage of the river had been attempted.
But finding their army lay quiet at Aberdeen, and that their advanced
party in Strathbogie only made now and then some little excursions
towards us as far as Keith, we, though at first we were very diligent
and alert, we relaxed in our vigilance, nay fell asleep and at last into
a lethargy, in which we unhappily continued till awakened by the foul
affair of Culloden, which merits an epithet bitterer than I can give.
About this time an advanced party under the command of Major
Glascow went out at night, and hearing that some of the enemy were
at Keith,[435] surrounded that place, and having taken most of them
prisoners, with their arms and equipage, returned the next day with
great honour to Spey-side, where Lord John Drummond and Lord
Ogilvy commanded.
 As we lay hereabouts a considerable time, assessments were
made upon the country for our support, and among others upon the
Earl of Findlater’s Estate, who on our arrival had taken wing and
joined Cumberland Will. The Earl’s Steward, being threatened with
military executions, if he refused to comply with our demands (which
were always seconded by such threats, though seldom put in
practice) begged leave to write to his Master for instructions how to
act. Upon this the Earl, having consulted with Cumberland, sent a
Letter addressed To the Man they call Lord John Drummond, telling
his Lordship, that if he or any other person should pretend to
exercise any military authority over any thing belonging to him, there
were Rebels’ houses enough, on which his Master Cumberland
promised him he should have his revenge. This Letter excited a
great deal of indignation among us; and was the cause of what
afterwards happened to his house;—for several of our party, without
any order being given, and indeed without the Prince’s being in the
then situation of his affairs, able to restrain their fury, ransacked it,
and carried away several articles of value, but without setting fire to
it, or wantonly destroying anything merely for destroying’s sake.[436]
At last news arrived that the enemy had left Aberdeen, and were
marching against us.[437] This intelligence gave great satisfaction to
many of us, who were in a manner tired out of our lives. Yet
notwithstanding the approach of the enemy, all the Prince’s
endeavours to collect his whole army, were ineffectual; for many
under pretence of cultivating their lands, or promising to come up
soon enough, went, staid, and came as they thought proper.
And now finding the enemy advanced pretty fast both by water
and land, the fleet bringing up their provisions and wearied soldiers;
frequent patrols were sent out by us in order to observe their
motions. Here I had the honour of commanding the last patrolling
party that ever crossed the Spey for the Prince’s cause. Lord John
Drummond (Commander at the Spey) having ordered me with ten
others to patrol all night towards the enemy, we began our work
about seven o’Clock and continued it till about five the next morning,
being then eight miles from Fochabers and two from the enemy;
when we took a man with a Letter from one of Cumberland’s
Secretaries to the Duchess of Gordon,[438] desiring her to employ all
her interest among her vassals in getting down provisions and
getting together what forces she could, as the Duke of Cumberland
intended to pass the river that day. Having secured the Messenger
and Letter, we continued our route, till we came up in a manner to
where they were encamped; for as they lay upon the declivity of a
hill, and had no guards on the top, we were able to approach very
near to them unperceived. But finding them drawn out in order of
battle, after seeing all we could see, and some bravadoes and
huzzas, we retired with all speed, leaving them to wonder what we
meant. We soon reached Fochabers (on the Spey) where I found
Lord John Drummond. Having given him the Letter taken from the
Duke of Cumberland’s Messenger and informed him of the situation I
had found the enemy in, I retired to repose myself a little while.
When fast asleep, a servant came in to tell me that the enemy was in
the town, and that it was too late to think of escaping, almost all of
our party having already passed the river. However, starting up in
great confusion, I resolved to risk all rather than fall into their hands,
and mounting my horse escaped by a back road. I had no sooner
crossed the river than I was ordered to join a party of about eighty
horse who were to remain behind on the banks of the Spey to
observe the motions of the enemy; whilst the foot, amounting to
about two thousand men, marched for Inverness, where the Prince
then lay with the greatest part of his army. Finding that the enemy
after a short halt at the Spey side, began to cross the river, we
likewise retired and followed our foot, to inform them of the enemy’s
being in full pursuit of us. This intelligence made our men pass
through Elgin, without halting, and straight on that night to Forres,
which was ten miles farther; and, after some stay there, to Nairn, out
of which we were next morning driven; the whole English Army
entering the town at one end, whilst we marched out at the other,
and continuing to pursue us sharply for three or four hours. And here
it was His Grace the Duke of Perth and Colonel O’Sullivan[439]
gained immortal honour by their bravery and conduct in bringing us
off in good order from under the very nose of the enemy; for
notwithstanding all their firing upon our rear, and though we were
much inferior in numbers, we lost not one man.
 Soon after their desisting from pursuing us, we received orders to
halt, and encamp upon that very place, where the fatal battle of
Culloden was afterwards fought. Having accordingly encamped as
well as we could on the heath that grew upon the common, which
served us both for bedding and fuel, the cold being very severe, we
were soon after joined by the Prince and several of his Clans.
Finding that the enemy did not pursue us we rested ourselves all that
night upon the Common, and early next morning drew out in battle-
array. But that day being Cumberland’s birthday,[440] and the enemy
shewing no intention to attack us on it, we reposed ourselves again,
though still keeping ourselves in readiness, upon the place where we
had rested the preceding night; a biscuit being given to us for our
refreshment. In that situation we remained, till the brave Prince came
amongst us in the dusk of the evening, with the full resolution of
going to attack them that night in their camp, distant only seven
miles. Orders were accordingly given to that effect, which were
obeyed with the greatest pleasure and alacrity by the whole army.
We began our March about seven o’Clock leaving great fires burning
in our camp: but by some strange infatuation or misfortune the road
was not rightly taken, either through the ignorance or treachery of
Lord George Murray’s guide. This still remains doubtful, but this I can
say, that with the little knowledge I had of the country I could have
conducted them much better and sooner. After we had marched till
about three o’Clock in the morning, over double the ground that was
necessary, we at last came pretty nigh the enemy’s camp: and when
we were supposing to surround them, and for that purpose in some
measure drawing out; my Lord George Murray began to be missing;
notwithstanding the Prince’s Aides-de-Camp in riding from rank to
rank, and asking, for God’s sake! what has become of His Lordship,
and telling that the Prince was in the utmost perplexity for want of
him. In that situation did we remain a considerable time, till, day
breaking fast in upon us, we heard that Lord George Murray was
gone off with most of the Clans. Where he had been all that
considerable time, or what was his intention in it, I leave Time to
prove. Now, after we had stood some time on the brink of entering
their camp, the Prince, on receiving the unwelcome news of Lord
George Murray’s going off with the greatest part of the army, was
under the necessity of ordering us likewise, much to our
dissatisfaction, to march back again to our Camp. In this manner did
that noble and well-concerted scheme fall to ruin—and not only to
ruin, but in such a manner as to ruin us, who before had hopes of
ruining the enemy. But O! for Madness! what can one think, or what
can one say here![441]
The enemy at our departure, being fully awakened and seeing
the jeopardy they had been in, judged rightly, it was their time now to
pursue us in the unhappy situation we were then in, being harassed,
hungry and starved and fatigued, almost to the greatest extremity.
We had no sooner reached our camp again, than news came of the
enemy’s being in full march towards us, and of their intention to
attack us. This disagreeable intelligence vexed us much in our
present situation, the more so, as of the twelve thousand men, who
were actually in arms and in the pay of the Prince, not above Four
thousand were now with him many going every moment,
notwithstanding his orders to the contrary, to Inverness, and to
woods and houses adjacent, in order to repose and rest themselves
after their late excessive fatigue. Many of these were so far from
rejoining us, that they were taken asleep by the enemy after the
battle. Those, however, who staid, put the best face on the affair they
could, and all of us presently appeared surprizingly courageous, who
only seemed to survive and animated by the spirit of loyalty and love
for our dear Prince. But now why we resolved to fight, or why we did
not retire to Inverness, and keep that town till we were fully joined,
which might have been easily done, or even at last, if judged proper,
avoided fighting and make another expedition into England in spite
of them,—I may say it was Fortune’s will;—for, contrary to the
Prince’s inclination, Lord George Murray insisted on standing and
fighting them that day: and as for what he said of our wanting
provisions, it is most certain, though we did that day, we might have
retired to Inverness and found there a sufficiency of meat for two or
three days. However, the Prince, notwithstanding his great inclination
to avoid fighting, was at last obliged to give way to the importunity of
Lord George Murray, who even used terms very cutting in case of
refusal; and was also for fighting His Grace the Duke of Perth—but
this may be said for him, he doubted not but the same Hand that had
supported and miraculously conducted the Prince hitherto would
infallibly continue to support him, and make him a glorious
Conqueror.[442]
I shall now proceed to give account in what manner we were
ranged in battle-array. The brave McDonalds, who till then had led
the van, and behaved at all times with great courage and bravery,
were now displaced, and made to give way, at the pleasure of Lord
George Murray, to the Athol men, whom he commanded. The rest of
the front line was composed of Highlanders: the second, of
Lowlanders and French, with four pieces of cannon at each wing:
and in the rear was the Prince attended by all the horse, and some
foot. In this manner were we drawn up—four thousand men to fight
eleven thousand. The enemy being by this time in full view, we
began to huzza and bravado them in their march upon us, who were
extended from right to left in battle-array, it being upon a common.
But, notwithstanding all our repeated shouts, we could not induce
them to return one: on the contrary, they continued proceeding, like a
deep sullen river; while the Prince’s army might be compared to a
streamlet running among stones, whose noise sufficiently shewed its
shallowness. The Prince, the Duke of Perth, the Earl of Kilmarnock,
Lord Ogilvy, and several other Highland and Lowland Chiefs, rode
from rank to rank, animating and encouraging the soldiers by well-
adapted harangues.
The battle being now begun, the whole fury of the enemy’s
Artillery seemed to be directed against us in the rear; as if they had
noticed where the Prince was. By the first cannon shot, his servant,
scarcely thirty yards behind him, was killed; which made some about
the Prince desire, that he would be pleased to retire a little off: but
this he refused to do, till seeing the imminent danger from the
number of balls that fell about him, he was by the earnest entreaties
of his friends forced to retire a little, attended only by Lord
Balmerino’s corps. Frequent looks and turns the Prince made, to see
how his men behaved: but alas! our hopes were very slender, from
the continual fire of musketry that was kept up upon them from right
to left. We had not proceeded far, when I was ordered back, lest the
sight of my standard going off, might induce others to follow. In
returning, various thoughts passed my soul, and filled by turns my
breast with grief for quitting my dear Prince, now hopes of victory,
then fear of losing—the miserable situation the poor loyalists would
again be reduced to—and what we had to expect if we left the field
alive: these thoughts, I say, strangely wrought upon me, till, coming
to the place I was on before, and seeing it covered with the dead
bodies of many of the Hussars who at the time of our leaving had
occupied it, I pressed on, resolving to kill or be killed. Some few
accompanied my standard, but soon left it. At this time, many of ours
from right to left were giving way and soon the battle appeared to be
irretrievably lost. The enemy, after we had almost passed the two
ranks, flanking and galling us with their continual fire, forced us at
last back, broke our first line, and attacked the second, where the
French troops were stationed. I happened then to be there, and after
receiving a slight grazing ball on my left arm, met with Lord John
Drummond, who, seeing me, desired I would come off with him,
telling me all was over and shewing me his regiment, just by him,
surrounded. Being quickly joined by about forty more horse, we left
the field of battle in a body, though pursued and fired upon for some
time. When we arrived at the foot of the hills, some of us took one
way, and some another: I, however, with about six more, continued
with Lord John Drummond; and it was with some difficulty we passed
the rapid torrents and frozen roads, till one o’Clock that night, when
we came to a little village at the foot of a great mountain, which we
had just crossed. Here we alighted, and some went to one house
and some to another. None of these cottages having the
conveniences to take in our horses, who wanted refreshment as well
as we, many of them perished at the doors. I happened to be in one
of the most miserable huts I had ever met with during my whole life;
the people were starving to death with hunger. However, having laid
myself down on the floor to rest myself after having been almost
thirty hours on horse-back; the people came crying about me and
speaking a language I did not understand, which made my case still
more unpleasant. But by good luck, a soldier soon after came in,
who could speak both to them and me, and brought with him some
meal, which was very acceptable, as I was almost starving with
hunger. Of this meal we made at that time a very agreeable dish, by
mixing it very thick with cold water, for we could get no warm: and so
betwixt eating and drinking we refreshed ourselves, till four o’Clock
in the morning; when Lord John Drummond and the rest of us began
our march, we knew not whither, through places it would be in vain to
describe; for we saw neither house, barn, tree, or beast nor any
beaten road, being commonly mid-leg deep in snow, till five o’Clock
that afternoon; when we found ourselves near a village called
Privana a Badanich,[443] the barracks of which, as I mentioned
before, the Prince had destroyed. Being now, to our surprize, almost
upon it, we consulted amongst ourselves how we might best get
intelligence from it; for, as it lay on the road from Inverness twenty-
four miles we apprehended the enemy might be there. But
fortunately a soldier coming out told us, that the village was occupied
by the Prince’s men. This intelligence gave us great pleasure; and
having accordingly entered the place, we found a great many of the
Prince’s adherents, the chief of whom was Lord George Murray and
the Duke of Perth; but we heard no news where the poor Prince was.
At first we had great hopes of rallying again: but they soon vanished,
orders coming for every one to make the best of his way he could.
So some went one way, some another: those who had French
Commissions surrendered; and their example was followed by my
Colonel, Lord Balmerino, tho’ he had none. Many went for the
mountains, all being uncertain what to do or whither to go.
In this perplexity I resolved to steer my course through the
mountainous country, notwithstanding the advice of many to the
contrary, who told me, it would be impossible for me to escape, and
begged I would go and surrender, assuring me, that if I attempted
the mountains, I should inevitably perish in them. But reflecting, how
nigh suffering my Father had been in the year 1715, taking Courage
and Patience for my guides, I resolved to enterprize a journey
through a Country that few of my Nation had ever passed before. So,
folding up my Standard, whose Motto was Britons! strike home! I put
it in my Riding-coat pocket, in hopes it might be of use another day,
and began my journey, in company with three others, for the
Highlands. Having discharged our horses, after a long day’s journey,
we came to a house situated on Garvie-more, twelve miles from any
other, where we met with many of our party, who had arrived there
before us. However, putting up in the best manner with what little we
could obtain, we set forward for Fort Augustus: but on the road, a
misfortune happened, that disconcerted all our plans; for a man who
carried our provisions of Oatmeal, fallen a little behind, by accident
met with some of the Brigade Picquets, who robbed him of our meal
and two riding-coats. This unexpected loss obliged us to separate
soon afterwards, being too many to subsist in this wild tract of
country, if we had kept together. However, having got betwixt Fort
Augustus and Fort William, we struck up into the country to the right,
and passed several large mountains in Lochiel’s Country, where we
staid three days, because we heard, that the brave Prince was nigh
us, and to take leave of one another, the necessaries of life being
exceedingly scarce, from the great number of people wandering over
the hills as well as we. I here went to wait upon the Duke of Perth,
who was at the house of Doctor Cameron, Lochiel’s Brother: but
being told by two sentinels at the door, that His Grace was
indisposed, I returned without seeing him. It was now reported, that
an English spy had been at Doctor Cameron’s house which obliged
me in all haste to quit that place; for certain it is, had I staid there any
longer, and the Highlanders supposed me to be the spy, they would
have made away with me. I therefore left my companions and set
forward to Lochaber, the wildest country I ever was in. And now it
was that I began to be truly miserable, and to endure hardships
which I had thought it impossible for human nature to support, for in
that most hideous place I was deprived of every thing that could give
me comfort: true it is, I found some few inhabitants; but in language
food and customs quite different from what I had ever seen before.
In this place I was forced to stay several days, on account of the
prodigious quantity of snow that fell upon the mountains, and
hindered me from discerning or making any road. During my stay, I
by good fortune got a pound of black bread to live upon. The snow
somewhat melting I set forward again from Lochaber towards the
sea-shore. On the road I was overtaken by about forty women, half-
starved to death who were wandering up and down for safety. Some
of them, who spoke English, told me, they had been driven out of
their houses by the soldiers who were sent out from Fort William to
ravage burn and plunder all before them and now it was that the
most heart-rending scenes of misery began to present themselves;
for many of these poor creatures with children in their arms, lay
extended in the clefts of the rocks half covered with snow, dead, and
a-dying in the most piteous manner. With these companions of
misery, and daily meeting with more I passed some days. I now
learned, that many of the unfortunate adherents of the Prince had
been famished to death on the hills and I expected it would soon be
my turn, for I began to be almost unable to proceed, my shoes being
worn out, and the sharp rocks wounding my feet. However, I
encouraged myself with the thought that my pursuers would have the
same difficulty to climb the rocks as I had; and on the twentieth
day[444] after our defeat at Culloden I came to the sea, in Clan-
Ronald’s Country; the view of which was most agreeable to me,
though even then I saw no prospect of escaping. Getting a little
refreshment from the people who dwelt on the sea-shore, I began as
it were to revive again, having been almost starved to death with
hunger and cold; for I had been obliged to lie down for whole nights
under the shelves of rocks, and was for two or three days together
without eating at all, as nothing could be obtained either for love or
money. Though I was fat and strong at the battle of Culloden, I was
now quite emaciated and reduced to so miserable a state, that, if I
had had another day to walk, I am sure I must have died; for I was
not only starved with hunger and cold, but frightfully covered with
vermin, which bit me all over my body so that there remained not
one whole place in my skin. This, joined with the pain in my torn feet,
made me often think that Job could not be in a more piteous
condition. Yet as he had God for his comforter, so had I; for the
justness of the cause I was suffering for, gave me great courage,
and supported me much: and though I saw daily enmities exercised
against me, it was a great satisfaction to me, to think, that, during the
time I had the honour of being a soldier under the banner of our dear
Prince, I could not accuse myself of one act that a Christian might
blush at.
 Being somewhat recovered by the particular care of a worthy
Gentleman (whose kindness I had the satisfaction in a little time to
return by an agreeable meeting with him at Paris), I began to inquire,
if it were possible from island to island to make my escape out of the
country; for could I have sold myself at that time as a slave into
Turkey, I would have done it. My host told me, that it was impossible,
as all the boats had been destroyed by Cumberland’s order.
However, one morning, being in that perplexity of thought how to get
off, and fearing every moment the landing of soldiers to destroy the
country—news was brought us that two French ships had come into
the Lough just by—which mightily raised our hopes, that either a
restoration of the Prince’s affairs were at hand, or that we should
escape to France. So, running down to see and hear what we could
we found them to be ships destined for the Prince’s service, having
on board a great quantity of arms and ammunition, with five barrels
of gold, pretty large and nearly one-yard long—which before our late
fatal disasters might have been of great use.[445]
Notice being sent all about the Mountains, as far as time would
permit; several, who lay despairing, came down to the sea-shore,
and among the rest, my old patron the Duke of Perth, Lord John his
brother; Sir Thomas Sheridan, Secretary Murray, Mr. John Hay, and
Doctor Cameron. These being assembled together, judged it proper
though no one knew where the Prince was (many thinking he was
gone off for France) to have the money and arms brought on shore;
which was done on the evening after. Going securely to sleep that
night expecting to sail for France the next day, we were surprised by
the noise of cannon, which awakened us about three o’Clock in the
morning; and getting up to see what the matter was, we had for our
comfort the disagreeable news and sight of three English ships, that
were come from Fort William to attack the French, whose
appearance on the coast they had noticed, it seems. This sight
displeased us very much: however, fighting was the resolution of us
all. The two French Frigates (viz., the Mars and the Bellona), being
pretty strong, and having a sufficient quantity of men, cannon, and
ball, resolved to make head against the three English vessels, of
which one was the Baltimore,[446] that name being written upon her
rudder, which was carried off by a cannon-ball. The place they fought
in, was a creek of rocks, which held the French (the English coming
down upon them) as it were penn’d up, having the land on their back
and both sides. However it was easy for them to hold
communications with us on shore, who were four hundred armed
men or more; so that had they been obliged to abandon their Ships,
they might have saved themselves on shore. The Crews of the two
ships amounting to nearly eleven hundred men, might, with the
assistance of the Highlanders, have made an effectual resistance to
the English, if they had attempted to invade us. The battle furiously
beginning at three o’Clock in the morning, it remained doubtful till
four in the afternoon, who would be victors. Nor was it a small
pleasure to us to see those combatants engaged, and the skill of the
French, whose fire seldom missed the English; for many of us being
upon the rock as it were hung over these ships, in such a manner
that they could not hurt us with either cannon or musketry; we could
discern how matters went, and few balls were fired but we might see
whether they hit or missed, which latter the English frequently did.
During the engagement, the Highlanders were busied in carrying the
arms, money, and powder off from the sea-shore; which service they
performed with amazing resolution, many a cannon ball being fired,
in order to hinder them, by the largest of the English ships. Few
Highlanders there were but what had a cask of brandy hid privately
in the hills, with which some of them got merry before night. At last
we had the satisfaction to see the English hoist their sails, leave the
French, and sail to the main ocean. The French repaired their ships
as fast as possible, and endeavoured to make what haste out they
could, lest the English should return with a greater force.
All being over and hopes reviving again; one who had been in the
Guards with me, came and told me, he had found a barrel of money,
and that he would get me as much of it as I pleased. To this proposal
I replied, That I had no manner of use for it, for, if I should be so
fortunate as to escape into France, I had friends enough there, who
would take care of me; and that if I died or were taken, it would be of
no service to me. Moreover if the Prince should rally again, how
shocking it would be to have to reproach ourselves with being a
hindrance to our dear Prince’s designs. On hearing this reply, he,