The Effect of Fines on Critical State and

Liquefaction Resistance Characteristics

of Non-Plastic Silty Sands Anthi
Papadopoulou & Theodora Tika
Visit to download the full and correct content document:
https://ebookmass.com/product/the-effect-of-fines-on-critical-state-and-liquefaction-re
sistance-characteristics-of-non-plastic-silty-sands-anthi-papadopoulou-theodora-tika/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...

Empirical Correlation of Soil Liquefaction Based on SPT

TV-Value and Fines Content Kohji Tokimatsu & Yoshiaki
Yoshimi

https://ebookmass.com/product/empirical-correlation-of-soil-
liquefaction-based-on-spt-tv-value-and-fines-content-kohji-
tokimatsu-yoshiaki-yoshimi/

Reformation, Resistance, and Reason of State

(1517-1625) Sarah Mortimer

https://ebookmass.com/product/reformation-resistance-and-reason-
of-state-1517-1625-sarah-mortimer/

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/

The Emergence of the Lyric Canon Theodora A

Hadjimichael

https://ebookmass.com/product/the-emergence-of-the-lyric-canon-
theodora-a-hadjimichael/
Memories of Resistance and the Holocaust on Film 1st
ed. Edition Mercedes Camino

https://ebookmass.com/product/memories-of-resistance-and-the-
holocaust-on-film-1st-ed-edition-mercedes-camino/

Effect of electrode Pt-loading and cathode flow-field

plate type on the degradation of PEMFC Qu

https://ebookmass.com/product/effect-of-electrode-pt-loading-and-
cathode-flow-field-plate-type-on-the-degradation-of-pemfc-qu/

Essentials of Plastic Surgery

https://ebookmass.com/product/essentials-of-plastic-surgery/

The Broken Sands (Empire of Usmad Book 1) Iren Adams

https://ebookmass.com/product/the-broken-sands-empire-of-usmad-
book-1-iren-adams/

Effect of protein on the post-ingestive response of

tilapia fed encapsulated diets Leandro S. Costa

https://ebookmass.com/product/effect-of-protein-on-the-post-
ingestive-response-of-tilapia-fed-encapsulated-diets-leandro-s-
costa/
 SOILS AND FOUNDATIONS Vol. 48, No. 5, 713–725, Oct. 2008
Japanese Geotechnical Society

THE EFFECT OF FINES ON CRITICAL STATE AND LIQUEFACTION

RESISTANCE CHARACTERISTICS OF NON-PLASTIC SILTY SANDS

ANTHI PAPADOPOULOUi) and THEODORA TIKAii)

ABSTRACT
Monotonic and cyclic triaxial tests were carried out on sand-silt mixtures for the investigation of the eŠect of ˆnes
content on their critical state and liquefaction resistance characteristics. Both the undrained and the drained monoton-
ic tests produce a unique critical state line for each tested mixture, which moves downwards with increasing ˆnes con-
tent up to a threshold value of 35z and then upwards. At a given void ratio and mean eŠective stress, the liquefaction
resistance ratio decreases with increasing ˆnes content up to the same threshold value of 35z, and increases thereafter
with further increasing ˆnes content. However, at a given intergranular void ratio, deˆned as the ratio of the volume of
ˆnes plus voids to that of sand particles, liquefaction resistance ratio increases with increasing ˆnes content up to the
threshold value. The threshold ˆnes content value, which is an important parameter in determining the transition from
the sand dominated to the silt dominated behaviour of sand-silt mixtures, is related to their particle packing. An ex-
pression is proposed for the estimation of the threshold ˆnes content as a function of the mean diameter ratio, d50/D50,
and the void ratio. The results, presented herein, also show that for each tested mixture the liquefaction resistance ratio
is related to the state parameter and that this relation is in‰uenced by the eŠective stress level and ˆnes content. The
results on the sand-silt mixtures are supported by similar results on natural silty sands.

Key words: critical state, ˆnes, liquefaction, non-plastic, sands, silt, state parameter, threshold, void ratio (IGC:
C3/D6/D7/E7)

modern codes, (NCEER, 1997 and Eurocode 8) and may

INTRODUCTION be explained by considering empirical correlations be-
Liquefaction of sandy soils under monotonic and cy- tween the relative density, Dr, and N value of the SPT,
clic loading conditions is considered to be one of the which also account for the grain size eŠects on the
major causes of failure of earth structures and founda- penetration resistance of sandy soils. Skempton (1986)
tions. Field observations from failures due to liquefac- found that the normalized (N1)60/Dr2 value increases with
tion have shown that both clean sands and sands contain- increasing grain size (or D50), which implies that for a
ing ˆnes (particles with diameter less than 75 mm) are sus- given (N1)60 value, Dr increases with decreasing D50 value
ceptible to liquefaction. Semi-empirical ˆeld-based (or increasing ˆnes content). Cubrinovski and Ishihara
procedures for evaluating the liquefaction potential dur- (2002) used the void ratio range, emax-emin to quantify the
ing earthquakes are based on correlations between ˆeld eŠect of grading properties on the SPT resistance. They
behaviour and in-situ index tests, such as standard found that the normalized (N1)60/Dr2 value decreases with
penetration test (SPT), cone penetration test (CPT), increasing emax-emin. As emax-emin increases with increasing
Becker penetration test (BPT) and shear wave velocity (V ˆnes content, the above ˆnding similarly indicates that
s). Seed et al. (1985) proposed the oldest and perhaps the for a given value of (N1)60, Dr increases as the ˆnes con-
most widely used procedure in which the cyclic stress ra- tent increases.
tio, CSR＝tav/s?o, is correlated with the SPT blow counts, In laboratory, although numerous studies have been
corrected for both eŠective overburden stress and energy, performed in order to investigate the eŠect of ˆnes con-
(N1)60, for clean sands and silty sands with ˆnes content tent on the liquefaction resistance ratio of silty sands, the
greater than 5z and earthquake magnitude, M＝7.5. Ac- results appear to be con‰icting. Ishihara (1996) and Ami-
cording to this correlation, the presence of ˆnes in silty ni and Qi (2000), among others, found that the liquefac-
sands increases their liquefaction resistance and conse- tion resistance ratio increases with increasing ˆnes con-
quently decreases the potential of liquefaction develop- tent (positive eŠect), and others like Troncoso and Ver-
ment. This beneˆcial eŠect of ˆnes on the liquefaction dugo (1985), Vaid (1994) and Miura et al. (1995) suggest-
resistance of silty sands has been also adopted in all ed that the liquefaction resistance ratio decreases with in-
i)
Research Student, Department of Civil Engineering, Aristotle University of Thessaloniki, Greece.
ii)
Professor, ditto (tika＠civil.auth.gr).
The manuscript for this paper was received for review on July 9, 2007; approved on July 23, 2008.
Written discussions on this paper should be submitted before May 1, 2009 to the Japanese Geotechnical Society, 4-38-2, Sengoku, Bunkyo-ku,
Tokyo 112-0011, Japan. Upon request the closing date may be extended one month.

713

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

714 PAPADOPOULOU AND TIKA

creasing ˆnes content (negative eŠect), whereas others
like Thevanayagam et al. (2000), Polito and Martin
(2001), Xenaki and Athanasopoulos (2003) and Yang et
al. (2004) found that there is an increase of liquefaction
resistance ratio with increasing ˆnes content up to a cer-
tain value and a decrease thereafter with further increas-
ing ˆnes content. The ˆnes content at which the eŠect of
ˆnes changes from positive to negative has been termed as
threshold, or limiting, or transitional.
The critical state, deˆned as the state at which the soil
continues to deform at constant shear stress and constant
void ratio, has increasingly been used as a fundamental
state to characterize the strength and deformation prop-
erties of sands in limit equilibrium (Casagrande, 1936;
Roscoe et al., 1958; Schoˆeld and Wroth, 1968). At this
state, there is a unique relationship among void ratio, ecs,
mean eŠective stress, p?cs, and shear strength, qcs, ex-
pressed by the critical state line, CSL, in the void ratio
versus mean eŠective stress plane. According to the criti-
cal state concept, the behaviour of a sand depends not
only on density, but also on stress level. The true state of
a sand is described by the location of its current state of
stress and volume relative to the critical state line. When
the state of a sand is above the critical state line, the sand
has the tendency to contract upon shearing, whereas
when its state is below the critical state line, it has a ten-
dency to dilate. Various normalized parameters have
been proposed to characterize the diŠerence between the
actual state and the critical state line. Been and JeŠeries
(1985) have quantiˆed the distance of the current state
from the critical state line by means of a state parameter,
c, which is the diŠerence in void ratios between the cur-
rent state and the critical state line at the current mean
eŠective stress, p?cs, Fig. 1.
Recently, in an attempt to re-evaluate the eŠect of
overburden stress considered in the semi-empirical ˆeld-
based procedures, such as SPT and CPT tests, the li-
quefaction resistance ratio of clean sands during both
monotonic and cyclic loading has been correlated with
the state parameter (Pillai and Muhunthan, 2001;
Boulanger, 2003; Idriss and Boulanger, 2004). The exten-
sion of the above stated correlation from clean sands to
sands containing ˆnes would be both useful and of great
interest.
The purpose of the work presented in this paper is ˆrst
to clarify the eŠect of ˆnes content, fc, on the critical state
line and the liquefaction resistance characteristics of silty
sands and then to investigate whether the liquefaction
resistance ratio of these soils can be related to their criti-
cal state.
TESTED MATERIALS
The materials used in the testing programme were ar-
tiˆcial sand-silt mixtures and natural silty sands.
Sand-silt mixtures were made from a clean quartz sand
(M31) with well-rounded grains and a non-plastic silt, a
ground product of natural quartz deposits from Assyros
in Greece. Samples were prepared by mixing the sand (S)
with the silt (F) at percentages of 5, 10, 15, 25, 35, 40 and
60z of the total dry mass of the mixture (noted as SF5,
SF10, SF15, SF25, SF35, SF40 and SF60 respectively).
Tests were also conducted on specimens of clean sand and
pure silt.
Two natural silty sands (D äuzce-1, D äuzce-2) were
retrieved from liqueˆed sites in D äuzce, Turkey, during
the 1999 earthquake of magnitude Mw＝7.1 and the other
was retrieved from the foundation of the Rimnio em-
bankment bridge in Greece, which failed during the 1995
Kozani earthquake of magnitude Ms＝6.6 (Tika and
Pitilakis, 1999).
The physical properties and grain size distributions of
the tested materials are presented in Table 1 and Fig. 2,
respectively.
Although the ASTM (D 4253 and D 4254) test methods
for the determination of minimum and maximum void
ratios are applicable to soils that may contain up to 15z,
by dry mass, of soil particles passing the No. 200 (75-mm)
sieve, provided they have cohesionless characteristics,
both these methods were used in this work in conjunction
with others in order to get a consistent value. In particu-
lar, for the determination of the minimum void ratio
both the vibratory table (ASTM D 4253, method 2A) and
the standard proctor test methods were used. The maxi-
mum void ratio was determined in accordance with the

Table 1. Physical properties of tested materials

Soils Gs emax emin D50 Cu fc

(mm) (z)*

Artiˆcial:
Sand (S) 2.649 0.841 0.582 0.30 1.3 0
Silt (F) 2.663 1.663 0.658 0.02 7.5 100
Natural:
D äuzce-1 2.732 0.724 0.366 0.38 9.7 13
D äuzce-2 2.741 1.155 0.521 0.19 12.5 23
Rimnio 2.667 1.059 0.419 0.10 5.3 37

Fig. 1. Deˆnition of state parameter (Been and JeŠeries, 1985) * (zº75 mm)
 TRIAXIAL TESTS FOR SAND-SILT MIXTURES
Fig. 2. Grain size distributions of the tested materials
Fig. 3. Variation of maximum, emax, minimum, emin, and critical state
(s?o＝100 kPa), ecs, void ratios with ˆnes content, fc, for the tested
materials
ASTM (D 4254, method C) test method and also by pour-
ing dry material in a mould three times and considering
the average as the maximum value. Figure 3 shows the
minimum and maximum void ratios of the tested materi-
als. As shown in the above Figure for the sand-silt mix-
tures, the minimum void ratio is obtained at a ˆnes con-
tent between 25z and 35z.
TESTING EQUIPMENT AND EXPERIMENTAL
PROCEDURE
The testing programme consisted of monotonic and cy-
clic triaxial tests for the determination of the critical state
line and the liquefaction resistance ratio of the tested
materials respectively. Both types of tests were performed
715
using a closed-loop automatic cyclic triaxial apparatus
(MTS Systems Corporation, U.S.A.). Its principles of
operation are given in detail in Papadopoulou (2008).
The specimens (height/diameter§100 mm/50 mm)
were formed by moist tamping at a water content varying
between 4z and 12z for all the tested materials and 35z
only for the silt specimens, using the undercompaction
method, introduced by Ladd (1978). Moist tamping was
preferred to other preparation methods, as it produces
specimens of varying densities (Verdugo and Ishihara,
1996). Saturation was achieved by percolating through-
out the specimen, from the bottom to the top drainage
line, ˆrst carbon dioxide gas (CO2) for 20 minutes and
then de-aired water. A suction pressure of 15 kPa was ap-
plied while dismantling the specimen, measuring its
dimensions and assembling the triaxial cell. In order to
ensure full saturation, a series of steps of simultaneous
increasing cell pressure and back pressure were per-
formed, while maintaining an eŠective conˆning stress of
15 kPa. In all the tests, a ˆnal back pressure of 500 kPa
was found to be su‹cient, as the parameter of pore water
pressure, B＝Du/Ds, did not increase by further increas-
ing back pressure. In all the tests the parameter B had
values from 0.95 to 1.00. No correction of the results was
made for membrane penetration, because of the uncer-
tainties, associated with such a correction and so the raw
test results are presented here. After the completion of
saturation, the specimens were isotropically consolidated
under an eŠective isotropic stress, s?o, ranging from 50 to
300 kPa. A period of time equal to double the consolida-
tion time of the specimens was allowed before shearing.
During consolidation the volume change and the axial
displacement of the specimens were recorded in order to
716 PAPADOPOULOU AND TIKA

calculate the post consolidation void ratio.

In the monotonic tests, the specimens were subjected to
either undrained (CU), or drained (CD) compression at a
constant rate of axial displacement of 0.05 mm/min.
In the cyclic triaxial tests, a sinusoidally varying axial
stress (±sd) was applied at a frequency of f＝0.1 Hz, un-
der undrained conditions. In this work, cyclic stress ratio,
CSR＝sd/2s?o, corresponds to double amplitude axial
strain, eDA§5z and liquefaction resistance or cyclic
resistance ratio, CRR15, is deˆned as the cyclic stress ra-
tio, CSR＝sd/2s?o, required to cause double amplitude
axial strain, eDA§5z at 15 cycles of loading.

TEST RESULTS AND DISCUSSION

Several studies have considered sands containing ˆnes
as consisting of two matrices, the sand grains matrix and
the ˆnes matrix, and analysed their behaviour in terms of
the interaction with each other (Vaid, 1994; Thevanaya-
gam et al., 2000; Polito and Martin, 2001; Xenaki and
Athanasopoulos, 2003; Yang et al., 2004). The nature of
the contribution of sand and ˆnes matrices may be ex-
pressed in terms of the intergranular and interˆne void
ratios respectively.
The intergranular void ratio, eg, expresses the relative
contribution of sand fraction on the behaviour of the
mixture and is given by the following equation (Mitchell,
1975):
VFINES＋VV fc＋w･(GSF/Sr) Fig. 4. Variation of (a) intergranular void ratio, eg (fcÃfcth) or (b) in-
e g＝ ＝ (1) terˆne, ef (fcÀfcth) void ratio with void ratio, e, at various fc (open
VSAND (1－fc)･(GSF/GSG) and closed symbols indicate emin and emax values respectively)
where VFINES is the volume of the ˆnes, VV is the volume
of voids, VSAND is the volume of sand grains, fc is the ˆnes
content, w is the water content of the specimen, GSF is the lowing the results for the critical state and the liquefac-
speciˆc gravity of the ˆnes and GSG is the speciˆc gravity tion resistance characteristics of the tested soils are
of sand grains. For saturated specimens (Sr＝100z) and presented and then discussed.
considering that GSF§GSG, the intergranular void ratio
after the consolidation of the specimen is expressed as Sand-silt Mixtures
follows: Critical State
fc＋ec In the monotonic undrained tests it was considered that
eg＝ (2) the onset of critical state conditions corresponds to the
(1－fc)
points where the shear and the mean eŠective stresses, as
where ec is the void ratio of the mixture after consolida- well as the pore water pressure, remained practically con-
tion. stant with axial strain. In the conducted tests, these
For mixtures with high ˆnes contents, when the sand points corresponded to strains ranging from 7z to 42z,
grains become isolated (VSAND may be ignored), the inter- depending on the type of behaviour (contractive, contrac-
ˆne void ratio, ef, may be a more appropriate parameter tive/dilative, dilative), the type of material and the den-
to be used (Thevanayagam and Mohan, 2000): sity. Figure 5 shows typical results from the undrained
tests in which contractive, contractive/dilative and dila-
VV
ef＝ (3) tive behaviour of specimens was observed. In the drained
VFINES tests it was considered that critical state conditions had
Similarly, for saturated specimens (Sr＝100z) and con- been reached at axial strains exceeding 30z, where the
sidering that GSF§GSG, the interˆne void ratio after the rate of change of both shear stress and volumetric strain
consolidation of the specimen is expressed as follows: decreased considerably or ceased. Figure 6 presents the
ec results from the drained tests. For all tested mixtures, the
ef＝ (4) CSLs in the q-p?cs plane are represented by straight lines
fc
passing through the origin:
Figure 4 presents the variation of both eg and ef with
qcs＝Mcs･p?cs (5)
the void ratio of mixtures at various fc values. In the fol-
 TRIAXIAL TESTS FOR SAND-SILT MIXTURES 717

Fig. 6. Results from monotonic drained triaxial tests, (a) q/p?c-ea and
(b) ev-ea plots

Table 2. Mcs values of tested materials

Soil Mcs R2
Fig. 5. Typical results of monotonic undrained triaxial tests with con-
tractive (SF35), contractive/dilative (S) and dilative behaviour Sand (S) 1.3553 0.9970
(SF15), (a) q-ea, (b) Du-ea and (c) q/p?c-p?/p?c plots SF5 1.4123 0.9991
SF10 1.4469 0.9997
SF15 1.5441 0.9848
SF25 1.4083 0.9999
where Mcs is an intrinsic constant for each mixture, Table SF35 1.4390 0.9994
2. An analytical description of the monotonic tests has SF40 1.4351 0.9999
SF60 1.4398 0.9992
been presented by Papadopoulou (2008).
Silt (F) 1.5147 0.9998
Figure 7 shows the CSLs in the e-p?cs of the sand-silt D äuzce-1 1.5013 0.9999
mixtures, as well as the sand and the silt. As shown in the Rimnio 1.3864 0.9996
above Figure, for each soil both the undrained and the
drained tests produce a unique CSL. At small mean eŠec-
tive stresses, below 300 kPa approximately, the CSLs are
nearly parallel and have a small inclination. With increas- coincide. The variation of ecs with fc, at an eŠective stress
ing mean eŠective stress level, however, they steepen and of s?o＝100 kPa, which is presented in Fig. 3, conˆrms
converge at stresses above 1000 kPa. Grading analyses the fcth value of 35z.
were performed on the initial materials, as used in the The CSLs of the sand-silt mixtures in terms of eg or ef,
tests and after the tests. No particle breakage was indicat- are shown in Fig. 8. As shown in this ˆgure, in this case
ed both for the sand and the sand-silt mixtures. As also the CSLs move upwards with increasing fc up to the fcth of
shown in Fig. 7, the CSLs move downwards with increas- 35z, and then downwards This may be attributed to the
ing fc up to a threshold value, fcth, of 35z, and then up- increasing values of eg as the increasing presence of silt
wards. The CSLs at fc values of 35z and 40z are very loosens the structure of the sand-silt mixture. At fc values
close, and the CSLs of the sand and the silt practically greater than the fcth＝35z, the presence of silt starts to
718 PAPADOPOULOU AND TIKA

Fig. 7. Critical state lines of sand-silt mixtures in terms of void ratio, e

Fig. 8. Critical state lines of sand-silt mixtures in terms of intergranu-
lar, eg ( fcÃ35%) or interˆne, ef ( fcÀ35%) void ratio

Fig. 9. Variation of cyclic stress ratio, sd/2s?o, with number of cycles, N, at constant void ratio, e, and various levels of mean eŠective stress, s?o,
for (a) S, (b) SF15, (c) SF25, (d) SF35 and (e) SF40
 TRIAXIAL TESTS FOR SAND-SILT MIXTURES
Fig. 10. Variation of cyclic stress ratio, sd/2s?o, with number of cy-
cles, N, at constant void ratio, e, and various ˆnes conents, fc, at
s?o, (a) 100 kPa and (b) 300 kPa
dictate the behaviour of the sand-silt mixture and ef is
used.
The above described behaviour for the sand-silt mix-
tures, concerning the inclination of the CSLs and the
eŠect of fc on their position, is in general agreement with
the results of the investigations of Zlatovi ác and Ishihara
(1995), Thevanayagam et al. (2002), Naeini and Baziar
(2001) and Yang et al. (2006). However, in most of the
above investigations, either the value of fcth was inferred,
or the inclination of the CSLs was determined over a nar-
row stress range.
Liquefaction Resistance
Figure 9 shows the variation of CSR with number of
cycles at various fc values. At a given number of cycles a
decrease of CSR with increasing s?o from 50 kPa to 300
kPa is observed for fc values up to 15z. Although,
results at s?o＝50 kPa and higher fc values are not availa-
ble, Figs. 9(c), (d) and (e) show that the eŠect of s?o on the
CSR diminishes as fc increases. The eŠect of fc on CSR at
constant void ratio and s?o is shown in Fig. 10. At a given
number of cycles, CSR ˆrst decreases with increasing fc
up to a threshold value of 35z and increases thereafter
with further increasing fc.
The eŠect of both s?o and void ratio on the CRR15 is
shown in Fig. 11. A decrease of CRR15 with increasing
void ratio at a given s?o is observed for each tested mix-
ture. The variation of the CRR15 with fc is shown in Fig.
719
Fig. 11. Variation of cyclic resistance ratio, CRR15, with void ratio, e,
for various ˆnes contents, fc, and at s?o, (a) 50 kPa, (b) 100 kPa and
(c) 300 kPa
12 at s?o＝100 and 300 kPa. The results in Figs. 11 and 12
show that once CRR15 reaches a minimum at a threshold
ˆnes content, it remains relatively constant thereafter
irrespectively of fc, unless the latter is increased sig-
niˆcantly. The threshold ˆnes content is about 35z and
25z at s?o＝100 and 300 kPa respectively. The threshold
ˆnes content at s?o＝100 kPa is identical to the fcth value,
determined from the monotonic tests, as described previ-
ously and shown in Figs. 3 and 7.
Figure 13 shows the variation of CRR15 with either eg
( fcÃ35z), or ef ( fcÀ35z) at a constant value of s?o. For
each tested mixture, it is shown that CRR15 decreases with
increasing eg or ef value. However, at a given eg value,
CRR15 increases with increasing fc up to the fcth＝35z,
(positive eŠect). At higher fc values ( fcÀ35z), a decrease
of CRR15 with increasing fc is indicated at a given ef value
(negative eŠect). This may be attributed to the fact that at
a constant value of eg, the void ratio of the mixtures
decreases with increasing fc and thus they become strong-
720 PAPADOPOULOU AND TIKA
Fig. 12. EŠect of ˆnes content, fc on liquefaction resistance ratio,
CRR15, at constant void ratio, e, at s?o, (a) 100 kPa and (b) 300 kPa
er (Fig. 4(a)). On the contrary, at a constant value of ef,
the void ratio of the mixtures increases with increasing fc
above the fcth value and they become weaker (Fig. 4(b)).
The positive eŠect of non-plastic ˆnes on CRR15 in
terms of eg is similar to that considered in the correlation
between CSR and (N1)60, in the SPT based procedure. As
noted previously, a given value of (N1)60, corresponds to
greater relative densities values, as fc increases, and there-
fore to greater values of CRR15. Similarly, for a given
value of eg, an increase in fc corresponds to an increase in
relative density and so in increased values of CRR15 (Fig.
4).
The Threshold Fines Content in Relation to Particle
Packing
As stated above, a change of the eŠect of ˆnes on both
the monotonic and cyclic behaviour of the tested mix-
tures was observed at fcth. This change of the eŠect of
ˆnes on the behaviour of mixtures can be explained by
considering the alteration of their structure with increas-
ing fc. As fc increases from zero, ˆnes enter the void
spaces between the sand grains, which are in full contact
with each other. This holds up to the fc at which the void
spaces are completely ˆlled with ˆnes, and results in an
increase of the density of the mixture, without however
any active participation of ˆnes in the transfer of inter-
particle contact stresses. With further increasing fc, the
sand grains start to separate, as they are pushed slightly
apart by the ˆnes, which start occupying also locations
between the sand grains and participating in the transfer
of interparticle contact stresses. This is the reason why a
decrease of ecs and CRR15 with increasing fc at a given
value of void ratio, is observed. This continues until the
Fig. 13. Variation of cyclic resistance ratio, CRR15, with intergranular
void ratio, eg (fcÃ35%) or interˆne, ef (fcÀ35%) void ratio
fcth, at which the volume of ˆnes is so great that sand
grains are displaced far apart from each other and they
loose contact. The behaviour of ˆnes now controls the
behaviour of the mixture, which becomes unstable. As fc
increases beyond fcth the sand grains become totally iso-
lated and the ˆnes matrix becomes stronger due to the in-
creasing development of contacts between ˆnes particles
which can now transfer larger interparticle contact stress-
es. This pattern of behaviour explains why for fcÀfcth an
increase of ecs and CRR15 with increasing fc at a given
value of void ratio, is observed.
Table 3 summarizes previous laboratory investigations
into the eŠect of fc, either on the critical state, or on the
cyclic behaviour of sand-silt mixtures in which fcth was de-
termined experimentally. According to them, fcth varies
from 35z to 50z. So far, there have been no studies on
how to estimate fcth on a theoretical basis. Considering
the packing of single size spherical sand particles with di-
ameter, D, at a separation distance, s, a relationship be-
 TRIAXIAL TESTS FOR SAND-SILT MIXTURES 721

Table 3. Threshold ˆnes content of silty sands

fcth fc
Sand Silt for n＝nth
Origin or Grain measured (eq. (8))
Ref. Soils Mineralogy shape d50/D50

emax emin D50 emax emin d50 fcth e fc

s?o kPa (z nth
(z) )

Toyoura sand & Quartz (77z) 30 1.50

(1) Feldspar (22z) SR-SA 0.977 0.600 0.170 1.754 0.62 0.010 0.059 À30** 0.500 100
milled Toyoura silt 35 2.00
Magnetite (3z)*
Quartz
(2a) Monterey No. 0/30 sand Feldspar & some
& Yatesville silt R-SR 0.821 0.631 0.430 1.723 0.727 0.030 0.070 35 0.680 100 35 2.30
Mica†

(2b) Yatesville sand & Natural SA-SR 0.972 0.653 0.180 1.723 0.727 0.030 0.167 37–50 0.760 100 37 1.15
50 1.72
Yatesville silt
Natural
(3) Sxinias-Marathon Quartz SA 1.040 0.660 0.120 1.710 0.660 0.020 0.167 44 0.650 200 44 1.28
sand & silt Feldspar &
Mica‡
Foudry 0S00 sand
(4) & crushed Quartz — 0.800 0.608 0.250 2.100 0.627 0.010 0.040 32.5!! 0.400 — 32.5 2.30
silica ˆnes

Ardebil sand & Natural

(5) Quartz & SR 1.090 0.746 0.190 1.720 0.765 0.025 0.132 35 0.780 100 35 1.40
Ardebil silt Feldspar

Sharp
Hokksund sand Quartz edges
(6) Feldspar & cubical 0.949 0.572 0.440 1.413 0.731 0.032 0.072 30 0.550 100 30 1.40
& Chengbei silt Mica
A-SA

(7) Sand (M31) & Quartz WR-SR 0.841 0.582 0.300 1.663 0.658 0.020 0.067 35 0.450 100 35 1.60
Assyros silt 0.650 2.30

Notation:
(1): Zlatovi ác and Ishihara (1995), (2a, 2b): Polito and Martin (2001), (3): Xenaki and Athanasopoulos (2003), (4): Thevanayagam et al. (2002),
(5): Naeini and Baziar (2004), (6): Yang et al. (2006a and b), (7): this study
WR: well rounded, R: rounded, SR: subrounded, SA: subangular, A: angular
* as reported by Verdugo and Ishihara (1996)
** the exact threshold ˆnes content was not determined
†
as implied from Polito and Martin (2001) and Mulilis et al. (1975)
‡
it was assumed, as most of the natural sands in Greece contain feldspar & mica and have subangular particles
!!
the exact threshold ˆnes content was not determined. It is estimated to be within the range of 25z to 40z

tween eg and the ratio of separation distance to particle

diameter s/D can be derived, Fig. 14:

Ø »
3
s
eg＝b･ ＋1 －1 (6)
D
where b＝1.910, 1.654 and 1.350 for the loosest (cubic),
average (triangular) and densest (tetrahedral) packing of
spherical grains. The separation distance represents the
thickness of the layer of ˆnes, which can be inserted be-
tween the spherical sand grains and indicates the
probability of a failure zone forming without the interfer-
ence of sand grains.
When mixing spherical particles of two diŠerent sizes,
the separation distance of the large size particles must be
n times the diameter of the small particles, that is:

Fig. 14. Relationship between intergranular void ratio, eg, and the ra-
s
D
＝ n･
d
D Ø » (7)
tio of the separation distance to the particle diameter, s/D, for sin-
gle size spherical particles where D and d are the diameters of the large and small
size particles respectively.
The formation of a shear zone at critical state would re-
722 PAPADOPOULOU AND TIKA

Fig. 15. Relationship between ˆnes content, fc, and ratio of diameters
of small to large size particles, d/D for the triangular packing (The
numbers in the ˆgure indicate the numbers of investigations, sum-
marized in Table 3)

quire the displacement of at least two to three diameters

of the small particles (JeŠeries and Been, 2006). Combin-
ing Eqs. (2), (6) and (7), the following expression of fc as
a function of diameter ratio, d/D, and void ratio can be
obtained:

Ø »
3
n･d
b･ ＋1 －(1＋ec)
D
fc＝ (8)
Ø »
3
n･d
b･ ＋1
D
Figure 15 shows the variation of fc with the diameter ra-
tio, d/D, for the average (triangular) packing, n＝1, 2
and 3 and void ratio values of 0.450, 0.650 and 0.760, as
well as the fcth values, determined by the investigations
summarized in Table 3. In Table 3 also the values of the Fig. 16. Variation of cyclic resistance ratio, CRR15, with state
parameter, c, for sand-silt mixtures at various levels of mean eŠec-
parameter n, denoted as nth, for which the fc values ob-
tive stress, s?o, (a) 50 kPa, (b) 100 kPa and (c) 300 kPa
tained from Eq. (8), matched the fcth determined by the
investigations, are presented. The mean diameter ratio,
d50/D50 and the same void ratio of the mixtures, as report- particles whose behaviour is dictated by the interacting
ed in the studies, were used to estimate fcth. For all the surface forces.
sand-silt mixtures, summarized in Table 3, the values of
the nth parameter range from 1.15 to 2.30. The above Relationship between State Parameter and Liquefac-
results show that the fcth depends mainly on the void ratio tion Resistance Ratio
of the mixtures, on the mean diameter ratio, d50/D50, as The value of CRR15 of the tested sand-silt mixtures is
well as the mineralogy and the particle shape characteris- related to the state parameter, c, in Fig. 16. For each
tics, as expressed by sphericity, angularity and roughness. tested mixture, a decrease of CRR15 with increasing value
Feldspar and mica minerals, encountered in many natural of c is observed, due to increasing contractiveness of the
silty sands (cases 1–3 and 5–6 in Table 3, nth＝1.15–2.30), soil. The test results also indicate that for dilative behav-
have platy shape and their presence facilitates the forma- iour (cº0) and low eŠective stresses (s?o＝50 kPa), the
tion of a shear zone at smaller separation distances be- presence of ˆnes up to 25z increases CRR15 for a given
tween the spherical particles (smaller n values), as com- c. This may be attributed to the fact that for fc up to 25z
pared with quartz sands with equidimensional particles, besides sand, silt particles also participate in transferring
(cases 4 and 7 in Table 3, nth＝1.60–2.30). or sustaining the imposed stresses. However, this contri-
It should be noted, however, that the grading of natur- bution of silt particles diminishes progressively with in-
al silty sands is in‰uenced by several other factors in addi- creasing s?o and at s?o＝300 kPa no apparent eŠect of fc on
tion to those considered above for binary packing and ar- the relation between CRR15 and c is observed for the test-
tiˆcial sand-silt mixtures. Natural sands have practically ed mixtures. There are also indications that for contrac-
an inˆnite number of particle diameters with varying tive behaviour (cÀ0), at s?o values equal or greater than
shape characteristics and may also contain very small 100 kPa, there is a lower bound value of CRR15 of the
 TRIAXIAL TESTS FOR SAND-SILT MIXTURES
Fig. 17. Critical state lines of natural silty sands in terms of (a) void
ratio, e and (b) intergranular void ratio, eg
Fig. 18. Variation of cyclic resistance ratio, CRR15, with (a) void ra-
tio, e and (b) intergranular void ratio, eg, at a mean eŠective stress,
s?o＝100 kPa
723
Fig. 19. Variation of cyclic resistance ratio, CRR15, with state
parameter, c, at the mean eŠective stress, s?o＝100 kPa
order of 0.09 to 0.12 independently of fc.
Natural Silty Sands
The CSLs of the natural silty sands are shown in Fig.
17 both in terms of void ratio and eg. The values of
parameter Mcs, determined for D äuzce-1 and Rimnio soils
are given in Table 2.
The variation of their CRR15 with both void ratio and
eg is shown in Fig. 18. The coincidence of CRR15 versus
void ratio relations between D äuzce-1 and the sand-silt
mixture with fc＝15z and also between D äuzce-2 and
Rimnio can be explained by the fact that the soils in both
cases had similar values of emin and emax, Table 1 and Fig.
3. Coincidence in the CRR15 versus eg relation is observed
only between D äuzce-1 and the sand-silt mixture with fc＝
15z, since both soils have also similar fc values.
724 PAPADOPOULOU AND TIKA
The relationship between CRR15 and c for natural soils
is shown in Fig. 19 together with the corresponding
results for the sand-silt mixtures with similar fc values. A
good agreement is observed between the results from
D äuzce-1 and the corresponding from the sand-silt mix-
ture with fc＝15z, as these soils have both similar emin
and emax and fc values. The agreement of the results be-
tween D äuzce-2 and the corresponding from the sand-silt
mixture with fc＝25z at cÀ0 may be regarded as ac-
cidental, since these soils have diŠerent physical proper-
ties (emin, emax and gradation). For Rimnio silty sand
higher values of CRR15 at a given value of c were deter-
mined than the corresponding for the sand-silt mixtures
with fc＝35z and 40z. This diŠerence may be attributed
to diŠerences in gradation, mineralogy, and particle
shape characteristics.
CONCLUSIONS
The following conclusions can be drawn from the test
results on the sand-silt mixtures:
Drained and undrained monotonic tests produce a
unique CSL for each tested mixture. However, the loca-
tion of the CSL is diŠerent for each mixture. At low
stresses, the CSLs are nearly parallel and have a small in-
clination. With increasing mean eŠective stress level,
however, they steepen and converge at stresses above
1000 kPa. The CSLs move downwards with increasing fc
up to a threshold value, fcth of 35z, and then upwards.
When eg or ef is considered, the CSLs move upwards with
increasing fc up to 35z and then downwards.
Liquefaction resistance ratio decreases with increasing
s?o at a given either void ratio, or eg (or ef) void ratio. The
eŠect of s?o on CRR15 diminishes with increasing fc. At
constant void ratio, CRR15 decreases with increasing fc up
to a fcth and increases thereafter with further increasing fc.
For the tested sand-silt mixtures the fcth is 35z at s?o＝100
kPa and is the same, as that determined from the CSLs.
The results also indicate that the fcth reduces with increas-
ing s?o. At constant eg, CRR15 increases with increasing fc
up to fcth (positive eŠect). This positive eŠect of ˆnes on
CRR15 in terms of eg is in agreement with that suggested
by the semi-empirical procedures in codes of practice,
when evaluating liquefaction potential during earth-
quakes.
The value of fcth is an important parameter, determin-
ing the transition from the sand-dominated to the silt-
dominated behaviour of the mixtures. It is shown that the
fcth of sand-silt mixtures is related to their particle pack-
ing. Equation (8) has been proposed for the estimation of
fcth, as a function of the void ratio of the mixture, the
mean diameter ratio, d50/D50, and the separation distance
of the sand grains which is equal to nth times the diameter
of silt particles with nth＝1.15–2.30. The comparison of
the fcth values, predicted by the above equation and the
corresponding determined experimentally by this and
previous studies, also highlights the in‰uence of other
parameters on fcth such as gradation, mineralogy and par-
ticles shape characteristics of the mixtures, which,
however, are not considered in the semi-empirical ˆeld
based procedures for the behaviour of natural silty sands.
The relation between CRR15 and c depends on both s?o
and fc. There are indications that the eŠect of fc diminish-
es with increasing s?o, i.e., at s?o＝300 kPa these relations
coincide for all mixtures, independently of fc and also
that for contractive behaviour (cÀ0) there is a lower
bound value for CRR15 of the order of 0.09 to 0.12 for the
tested mixtures. The behaviour of natural soils is similar
to that of artiˆcial mixtures, only when they have similar
physical properties, such as gradation, mineralogy and
particle characteristics.
It is worth noting, however, that the in-situ relation-
ship between liquefaction resistance ratio and state
parameter may be diŠerent from that determined in
laboratory due to the eŠects of stress history, structure
(fabric and bonding) and ageing.
ACKNOWLEDGEMENTS
The authors would like to express their acknowledge-
ments to the Onassis Public Beneˆt Foundation, the State
Scholarships Foundation of Greece and the General
Secretary of Research and Technology, Greece, for the
ˆnancial support of the ˆrst author. The tests on the soils
from D äuzce were performed within the Marmara Earth-
quake Rehabilitation Project (MERP).
REFERENCES
1) ASTM D 4253: Standard Test Methods for Maximum Index Den-
sity of Soils Using a Vibratory Table, American Society of Testing
and Materials, U.S.A.
2) ASTM D 4254: Standard Test Methods for Minimum Index Density
and Unit Weight of Soils and Calculation of Relative Density,
American Society of Testing and Materials, U.S.A.
3) Amini, F. and Qi, Z. G. (2000): Liquefaction testing of stratiˆed sil-
ty sands, J. Geotech. and Geoenvr. Engrg., ASCE, 126(3),
208–217.
4) Been, K. and JeŠeries, M. G. (1985): A state parameter for sands,
G áeotechnique, 41(3), 365–381.
5) Boulanger, R. W. (2003): High overburden stress eŠects in liquefac-
tion analyses, J. Geotech. and Geoenvr. Engrg., ASCE, 129(12),
1071–1082.
6) Casagrande, A. (1936): Characteristics of cohesionless soil aŠecting
the stability of slopes and earth ˆlls, J. Boston Soc. Civil Engrg.,
13–32.
7) Cubrinonski, M. and Ishihara, K. (2002): Maximum and minimum
void ratio characteristics of sands, Soils and Foundations, 42(6),
65–78.
8) Eurocode 8-EN 1998: Design of Structures for Earthquake
Resistance, Part 5: Foundations, retaining structures and geo-
technical aspects.
9) Idriss, I. M. and Boulanger, R. W. (2004): Semi-empirical proce-
dures for evaluating liquefaction potential during earthquakes,
Proc. 11th Int. Conf. on Soil Dyn. and Earth. Engrg.-3rd Int.
Conf. on Earth. Geotech. Engrg.
10) Ishihara, K. (1996): Soil Behaviour in Earthquake Geotechnics,
Clarendon Press Oxford, New York, U.S.A., 227.
11) JeŠeries, M. and Been, K. (2006): Soil Liquefaction-A Critical State
Approach, Taylor and Francis, London.
12) Ladd, R. S. (1978): Preparing test specimens using undercompac-
tion, Geotech. Test. J., GTJODJ, 1(1), 16–23.
13) Mitchell, J. K. (1975): Fundamentals of soil behavior, John Wiley
 TRIAXIAL TESTS FOR SAND-SILT MIXTURES
and Sons, Inc., New York, U.S.A., 1st Edition.
14) Miura, S., Yagi, K. and Kawamura, S. (1995): Liquefaction
damage of sandy and volcanic grounds in the 1993 Hokkaido Nan-
sei-Oki earthquake, Proc. 3rd Int. Conf. on Rec. Adv. in Geotech.
Earth. Engrg. and Soil Dyn., 3.06, 193–196.
15) Mulilis, J. P., Chan, C. K. and Seed, H. B. (1975): The eŠects of
method of sample preparation on the cyclic stress-strain behaviour
of sands, EERC Report, (75–18).
16) Naeini, S. A. and Baziar, M. H. (2004): EŠect of ˆnes content on
steady-state strength of mixed and layered samples of a sand, Soil
Dyn. and Earth. Engrg., 24(3), 181–187.
17) National Center for Earthquake and Engineering Research
(NCEER) (1997): Proc. NCEER Workshop on Evaluation of Li-
quefaction Resistance of Soils, Report No. NCEER–97–0022,
Youd, L. T. and Idriss, I. M, 6, 62, 64.
18) Papadopoulou, A. I. (2008): Laboratory investigation into the be-
haviour of silty sands under monotonic and cyclic loading, Ph.D
Thesis, Aristotle University of Thessaloniki, Greece.
19) Pillai, V. S. and Muhunthan, B. (2001): A review of the in‰uence of
initial static shear (Ka) and conˆning stress (Ks) on failure mechan-
isms and earthquake liquefaction of soils, Proc. 4th Int. Conf. on
Rec. Advn. Geotech. Earth. Engrg. and Soil Dyn., San Diego,
California, 1.51.
20) Polito, C. P. and Martin, J. R. (2001): EŠects of nonplastic ˆnes on
the liquefaction resistance of sands, J. Geotech. and Geoenvir.
Engrg., ASCE, 127(5), 408–415.
21) Roscoe, K. H., Schoˆeld, A. N. and Wroth, C. P. (1958): On the
yielding of soils, G áeotechnique, 8(1), 22–53.
22) Schoˆeld, A. N. and Wroth, C. P. (1968): Critical State Soil
Mechanics, Mc-Graw Hill, London.
23) Seed, H. B., Tokimatsu, K., Harder, L. F. and Chung, R. M.
(1985): The in‰uence of SPT procedures in soil liquefaction
resistance evaluations, J. Geotech. Engrg., 111(12), 1425–1445.
24) Skempton, A. K. (1986): Standard penetration test procedures and
the eŠects in sands of overburden pressure, relative density, particle
725
size, aging and overconsolidation, G áeotechnique, 36(3), 425–447.
25) Thevanayagam, S., Fiorillo, M. and Liang, J. (2000): EŠect of non-
plastic ˆnes on undrained cyclic strength of silty sands, Soil Dy-
namics and Liquefaction, 77–91.
26) Thevanayagam, S. and Mohan, S. (2000): Intergranular state varia-
bles and stress-strain behaviour of silty sands, G áeotechnique, 50(1),
1–23.
27) Thevanayagam, S., Shenthan, T., Mohan, S. and Liang, J. (2002):
Undrained fragility of clean sands, silty sands and sandy silts, J.
Geotech and Geoenvir. Engrg., ASCE, 128(10), 849–859.
28) Tika, T. and Pitilakis, K. (1999): Performance of Rimnio bridge
embankment during the 1995 Kozani-Grevena earthquake, Proc.
12th European Conf. on SMGE, The Netherlands, 2, 857–862.
29) Troncoso, J. H. and Verdugo, R. (1985): Silt content and dynamic
behavior of tailing sands, Proc. 12th ICSMFE, San Francisco,
U.S.A., 1311–1314.
30) Vaid, V. P. (1994): Liquefaction of silty soils, Ground Failures un-
der Seismic Conditions, Geotechnical Special Publication, (44),
ASCE, 1–16.
31) Verdugo, R. and Ishihara, K. (1996): The steady state of sandy
soils, Soils and Foundations, 36(2), 81–91.
32) Xenaki, V. C. and Athanasopoulos, G. A. (2003): Liquefaction
resistance of sand-silt mixtures: an experimental investigation of
the eŠect of ˆnes, Soil Dynamics and Earth. Engrg., 23, 183–194.
33) Yang, S., Sandven, R. and Grande, L. (2004): Instability of loose
sand-silt mixtures, Proc. 11th Int. Conf. on Soil Dynamics and
Earth. Engrg.-3rd Int. Conf. on Earth. Geotech. Engrg., Berkeley,
California, U.S.A.
34) Yang, S., Sandven, R. and Grande, L. (2006): Steady-state lines of
sand-silt mixtures, Can. Geotech. J., 43, 1213–1219.
35) Zlatovi ác, S. and Ishihara, K. (1995): On the in‰uence of non plastic
ˆnes on residual strength, Proc. 1st Int. Conf. on Earth. Geotech.
Engrg., 14–16 November 1995 (eds. by K. Ishihara, A. A. Balke-
ma), Rotterdam, The Netherlands, 239–244.
Another random document with
no related content on Scribd:
The Project Gutenberg eBook of Pyramids of snow
This ebook is for the use of anyone anywhere in the United States
and most other parts of the world at no cost and with almost no
restrictions whatsoever. You may copy it, give it away or re-use it
under the terms of the Project Gutenberg License included with this
ebook or online at www.gutenberg.org. If you are not located in the
United States, you will have to check the laws of the country where
you are located before using this eBook.

Title: Pyramids of snow

Author: Edith Metcalfe

Release date: December 16, 2023 [eBook #72428]

Language: English

Original publication: London: Ward, Lock & Co., Limited, 1903

Credits: Al Haines

*** START OF THE PROJECT GUTENBERG EBOOK PYRAMIDS

OF SNOW ***
"'Go out and stop out, or I'll have you put out.'" (Page 83.)
 PYRAMIDS OF SNOW.

BY

EDITH METCALFE.

"I can tell you without the help of an augur what will be your fate you become a
gambler. Either the vice will end by swallowing you up alive as a quicksand does, or if you
are a winner, your gains will disappear more quickly than they came, melting like pyramids
of snow."

WILLIAM DE BRITAINE.

LONDON:
WARD, LOCK & CO., LIMITED.
NEW YORK AND MELBOURNE.
1903

CONTENTS.
CHAPTER I
The Viaticum

CHAPTER II.
The Best Thing in the World

CHAPTER III.
Fraud

CHAPTER IV.
Mediation

CHAPTER V.
Kindred and Affinity

CHAPTER VI.
Bravado

CHAPTER VII.
Melville leads Trumps

CHAPTER VIII.
Rivals

CHAPTER IX.
Bigamy

CHAPTER X.
Light come, Light go

CHAPTER XI.
Mrs. Sinclair pays a Visit

CHAPTER XII.
A Pic-nic
CHAPTER XIII.
Murder

CHAPTER XIV.
The Finding of the Body

CHAPTER XV.
Flight

CHAPTER XVI.
An Unexpected Will

CHAPTER XVII.
An Arrest

CHAPTER XVIII.
A Faithful Servant

CHAPTER XIX.
In the Park

CHAPTER XX.
Money makes a Difference

CHAPTER XXI.
The Result of the Trial

CHAPTER XXII.
Mr. Tracy becomes Active

CHAPTER XXIII.
Sir Ross is Quits

CHAPTER XXIV.
Mrs. Sinclair Resolves to Go Away

CHAPTER XXV.
Mrs. Sinclair Goes away
CHAPTER XXVI.
Fate takes the Odd Trick

CHAPTER XXVII.
The Place of Peace

PYRAMIDS OF SNOW.

CHAPTER I.

THE VIATICUM.

Upon most of the people who thronged the rooms the incident was lost.
Of those who saw it many did not understand its meaning, and the rest were
too much absorbed in their own affairs to give it any attention. The scene
was the Casino at Monto Carlo; every chair was occupied, and behind every
chair men and women were standing, all intent upon the play, all consumed
by the feverish thirst of winning money born of the atmosphere of the place.
The brilliant light flashed in jewels and gleamed in eager eyes, heightened
the colour of flushed cheeks and emphasised the pallor of haggard faces;
against the black evening coat of one man sitting down was outlined the
bare arm of a woman, who laid her stake upon the table, and when the hand
was withdrawn it still hesitated over the black coat until the fortune of the
stake should be declared. Dominating everything was the monotonous
sound of the croupiers' voices and the noise of the money as it was raked to
and fro upon the tables.
 The incident which took place in this scene was a not uncommon one. It
was a little procession of three men, one a dark, good-looking man in well-
cut evening dress, who walked nonchalantly through the rooms, pausing
almost imperceptibly while his two companions shot a glance of
interrogation at each of the croupiers; when the croupiers, in reply, had shot
a glance of assent at his companions, the dark man moved on again until he
had almost completed his tour of the rooms. It was Melville Ashley
undergoing the process of identification as a well-known frequenter of the
rooms before receiving the viaticum which should enable him to return to
London.

It is the habit of the Englishman to conceal his feelings, and no one

could have guessed from Melville's demeanour whether he experienced
relief at having come to the end of his tether, regret at knowing that he
could play no more that season, mortification at his somewhat humiliating
position, or any other emotion which one may suppose natural to a gambler
who is suddenly baulked in his pursuits. He seemed entirely unconcerned,
perhaps a little bored, but certainly in complete possession of himself. To
the few people who, knowing him, found time to vouchsafe him a nod of
greeting, he bowed pleasantly enough. Of the existence of the others he
appeared unaware, though, in point of fact, his senses were so alert that he
could have supplied a remarkably close description of everyone had he been
asked to do so. For the time the gambling fever had left him, and with the
vanishing of his last coin there awoke in his mind an intense disgust at the
heavy scent in the air and the grotesque sight of the many pairs of white
gloves. He was only anxious for the great baize doors to swing behind him
and exclude him from what was generally the one desire of his heart.

Only once did he betray any interest. A woman leaning back in her
chair put out her hand to detain him. She understood the significance of his
escort, and there was some commiseration in her eyes.

"Are you going home, Mr. Ashley?" she asked, in a low tone.

"Yes," he answered, with a little smile; "I leave to-night."

In those conventional words was conveyed a perfectly frank confession

of the state of his finances. No need to invent any explanation of so sudden
a departure. His questioner was well enough acquainted with the language
of the place to know that he had pledged his word to return at once to
London, in consideration of value received.

"I'm sorry," she said, and looked as if she meant it; "but I daresay I shall
be following you soon, and then, perhaps, we may meet again. London is a
tiny little place."

"Yes," Melville assented politely; "but wouldn't it be as well if you gave

me one of your cards?"

"I haven't any," said Mrs. Sinclair, smiling lightly, for she liked a
sportsmanlike loser. "Men always carry cards—in case of duels, I suppose,
but women have no room in their purses for anything but money, and
nowhere but their purses to put anything else. Give me one of yours, and I
will write to you."

"That is too good of you," he replied, as he gave her one; "but of course
you will forget all about it. Good-night, good-bye."

"Auf wiedersehen," she answered prettily, and turned to her companion

on her left, who had watched the little comedy with a scowl upon his face.
Melville noted the scowl and bowed sardonically as he moved away. To be
conscious of superiority to anyone is satisfactory in one's hour of
discomfiture, and Melville derived a complacent satisfaction from this little
man's evident annoyance.

"The little bounder doesn't like me," he thought, "but he's a little ass to
show it. He must be very rich for Mrs. Sinclair to be willing to lay aside her
weeds for him."

The doors swung behind him, and in another moment Melville was in
the open air. He stretched out his arms in pure enjoyment of the lovely
night.

"I am infinitely obliged to you," he said to his escort; "the other trifling
formalities will, doubtless, be completed in due course;" and in what
seemed an incredibly short time Melville was on his way to London.
 Inside the Casino, the little bounder turned to his companion.

"Since you have no room in your purse for visiting cards," he said, "may
I not keep that one in safe custody for you?"

"Thanks, no," the woman answered, and slipped it inside her dress; "I
haven't finished playing yet, and my luck is in to-night."

"Would you be as kind to me," he pursued, "if I had to have recourse to

the charity of the bank to pay my fare to London? Or would you drop me
when my money went?"

Mrs. Sinclair looked at him coolly.

"Don't ask leading questions, and please don't make yourself ridiculous.
Civility costs nothing, and it amused me to be civil to that—gentleman."

"It is rare for you to be amused with anything that costs nothing," he
retorted, but Mrs. Sinclair would not be drawn. She began to play again,
and, when at last she stopped, the little man's carrying capacity was taxed to
take her winnings back to her hotel.

It would be a vain task to try to record all Melville Ashley's thoughts as

the train bore him across France; in the aggregate they amounted to little
less than a comprehensive cursing of everything and everybody, including
himself. For his position was desperate.

The younger son of his parents, both of whom had died while he was
still an infant, he had been brought up with his brother Ralph under the
guardianship of his uncle, Sir Geoffrey Holt, lord of the manor of
Fairbridge, in Surrey, whose co-heir, at any rate, he hoped some day to be.
Sir Geoffrey had played his part well, placing every advantage in the way of
both his nephews, but as the years slipped by he found it difficult to be quite
impartial in his personal treatment of the two lads, though he never failed to
be impartial in his dealings with them so far as they affected the education
and up-bringing of the boys.
 It was Ralph, however, who engrossed his uncle's affection, and
something in Melville's nature rose in rebellion at the thought that he came
second in the estimation of any person. Both boys were handsome, Melville
especially so; both were well endowed with intelligence, and both took
advantage of their opportunities. But whereas Ralph developed into a frank
and unaffected man, fond of athletics and outdoor pursuits, Melville became
more and more self-centred and reserved, devoting all his time to his one
absorbing love of music. Manhood brought liberty, and liberty in Melville's
case brought lack of self-restraint. His finer qualities led him into a certain
sort of temptation, and the men with whom his rare musical talents brought
him into contact were of a free and easy Bohemian type that did not afford
the most healthy companionship for a young fellow of his particular
temperament. Musical evenings led to smoking concerts, and the concerts
to late nights of which other and less innocent amusements were the
principal feature; billiards and cards became first a habit and then a passion,
and Melville was still in his early twenties when it was obvious that he was
a confirmed gambler.

Sir Geoffrey was patient and he was rich, but detestation of the gambler
was added to his dislike of his younger nephew, and more than one violent
quarrel had taken place between the two. It says much for the elder man that
he never referred to the position of absolute dependence occupied by the
younger one; but when, a few weeks before, Melville came to him with the
oft-repeated tale, Sir Geoffrey spoke his mind in the vernacular.

"Let me know the sum total of your accursed debts," he said, "if you
have the honesty or the wit to remember them, and I will clean the slate.
Then I will give you a final two hundred and fifty for yourself, and that
shall be the end."

When Melville gave him the damning list of debts, Sir Geoffrey bit his
lips until they bled. Livery stables, and wine and cigar merchants told a tale
of luxurious living which Sir Geoffrey himself had never been able to
afford in his younger days, and there were other items not precisely
specified, into which the elder man thought it better not to enquire too
curiously. But he kept his word. He drew crossed cheques payable to every
person named in the list for the full amount, and demanded a receipt from
each in full discharge of his nephew's liability. When the last receipt came
in, after a miserable week of waiting, he sent for Melville to his library.

"Is that the last?" he enquired grimly, and Melville assented. Then Sir
Geoffrey sat down at his table and drew one cheque more. "There is the two
hundred and fifty I promised you," he said; "make the best use of it you can,
for it is the last you ever have from me. The dog-cart will take you to the
station in half-an-hour." Then he turned on his heel and left him, and
Melville returned to town.

Five weeks before! And now the whole of the money was gone. With all
his ingenuity it would be difficult to invent a story which his uncle would
be likely to accept as a valid explanation of so surprising a fact.

Melville lighted a cigar and cursed his luck again.

Then the gambler's spirit re-asserted itself. He had had a glorious time at
Monte Carlo while it lasted. One night he had won more than five thousand
pounds, and another night the bank had to send out twice for fresh supplies
of money. That was the time of triumph. People had crowded round him,
some to follow his play, some to envy, some to congratulate him, and
among them he had seen Lavender Sinclair for the first time: a magnificent
woman truly, with splendid colouring and grandly moulded limbs; she wore
turquoise velvet, he remembered, and round her neck a barbaric collar of
turquoise bosses linked together on red gold; even in that room, where
jewels were as common as morals were rare, her jewels were conspicuous,
and she wore them perfectly. Some acquaintance introduced him to her, and
she seemed interested in hearing his name—had met people who knew him,
or some distant kinsmen, but there was no indication of any desire on her
part to press the acquaintance. She was in the ripest glory of her beauty, the
sort that is at its best when it is mature. He wondered idly how old she was,
over thirty certainly; but, after all, it did not matter. Rumour had it that she
was going to marry Sir Ross Buchanan, and Melville was contemptuous of
her choice of a second husband; he knew the man by sight, an undersized,
rather weakly fellow, who inherited an old title from his father and, it was
said, two millions sterling from his mother. Sir Ross was a pill that required
an unusual amount of gilding, and Melville's first admiration of the woman
was replaced by scorn of her venality. She was sympathetic though when he
bade her good-bye, and Melville appreciated sympathy.

The journey was very tedious, so Melville opened his dressing-case and
took out a packet of letters which had reached him at the hotel, but to which
he had not troubled to attend. Several he tore up and threw away, but there
was one which he carefully replaced in its envelope in his bag. It was from
his brother, and ran as follows:

"DEAR MELVILLE,—Why didn't you tell me you were going to

Monte Carlo? However, I hope you are enjoying yourself and having good
luck. By the way, I am going to ask you to do me a great favour. Can you
lend me a hundred for a fortnight? I will repay you then. My solicitors are
selling some capital for me, but they are so slow, and I am in immediate
want of the money. Do write soon.—Yours ever, RALPH ASHLEY. P.S.—
Have you heard of my engagement to Gwendolen Austen?"

"So he is hard up, too," Melville muttered. "No, I wouldn't lend him
fifty pounds if I had fifty thousand to-morrow. And engaged to Gwendolen
is he? I wonder if I can put an end to that. If she were my wife I might even
win the old man round again."

Then his mind reverted to his immediate difficulties, and he went over
the old useless ground of trying to think of some way to raise the wind,
failing once more to see any light at all, as indeed he was bound to fail,
since honest work did not come into his most casual consideration.

It was not, however, until he found himself in his chambers in Jermyn

Street that he fully realised how he had come to the end of all things. There
were invitations awaiting him which he could not accept for lack of ready
money; little accounts which he would have been only too glad to hand over
to his uncle if he had remembered their existence; all insignificant enough
individually, but totalling up to a considerable sum; private tips from
hangers-on at stables, which were certain to be good since he could not
avail himself of them; letters from women suggesting trips up the river or
supper after the play; even letters from friends saying they were hard up,
and reminding him of small obligations under which he lay to them.
Melville felt as if he were at last at bay, with all his worries like so many
starving wolves tearing him down to his destruction. And worse than all
was the extreme physical reaction from the unwholesome life of excitement
he had lately been leading at Monte Carlo. While that life lasted no fatigue
oppressed him. A tumbler of champagne or a stiff pick-me-up from a
chemist always availed to keep him going. But now the excitement was
over. The curtain was rung down, the lights were all turned out, he was
alone with his troubles, and had no pluck left to face them. In sheer
weariness he turned into bed and slept the sleep of deep exhaustion.

CHAPTER II.

THE BEST THING IN THE WORLD.

Even while Melville, with despair gnawing at his heart, was speeding on
his journey back to England, Sir Geoffrey Holt was keeping festival at
Fairbridge Manor. That very evening he had given a final dinner party to
celebrate the betrothal of his god-child, Gwendolen Austen, to his favourite
nephew, Ralph Ashley.

In the whole of a land which is proud to claim as its children so many

fair women and brave men, it would be difficult to find a fairer woman or a
braver man than now engrossed Sir Geoffrey's thoughts, and in their
approaching union he looked to see the culmination of his own happiness. It
was infinitely pleasant to know that the two, over whose lives he had
watched so tenderly, would never leave him now, but hand-in-hand would
walk in quiet contentment by his side, lightening the burden of his
increasing years, and giving him fresh pleasure in their own unfolding joys.
No man could ever hope to win richer reward for his unfailing goodness to
others than Sir Geoffrey was reaping now for his long care of this boy and
girl.

So he threw wide his hospitable doors, and asked the county to come
and shower congratulations upon the happy couple. For a week he kept
open house, and his pleasure was so apparent, his high spirits so contagious,
that he made himself loved the more by his unaffected delight and his
manner of displaying it. To his succession of dinner parties practically the
entire county came, until both Ralph and Gwendolen were at a loss to find
fresh ways of saying, "Thank you," for so many expressions of goodwill.

But this evening had brought the entertainments to a close, and when Sir
Geoffrey, standing by his open door, had bade the latest guest good-bye, he
turned with a sigh of satisfaction into the great hall where his children, as he
called them, were laughing over some incident which had amused them
during the day.

Sir Geoffrey pulled his god-daughter towards him and held her face
between his hands.

"The last guest gone," he said, smiling at her; "now, Gwen, confess you
are not sorry."

"I didn't know there was so much kindness in the world," she answered,
smiling back at him, and her eyes were shining; "but I confess I am glad we
are all by ourselves again."

"Tired?" he asked.

"Not a bit," she answered brightly; "unless it be of seeming to occupy so

much attention."

"And you don't want to go to bed?"

"Indeed, no," she said indignantly. "When one is as happy as I am it

would be a shame to spend a single hour asleep."
 "Then let us go down to the house-boat," Sir Geoffrey said. "I daresay
Ralph can manage to amuse you somehow, and I want to talk to your
mother."

"Do you want to talk to me, Ralph?" said Gwendolen, turning to her
lover, who was looking at her with affectionate pride.

"I don't seem to have had a chance of talking to you for a week," Ralph
answered promptly. "Let's go at once and—and get a deck chair ready for
your mother."

Sir Geoffrey chuckled.

"An admirable reason for both of you hurrying away. Ralph is too weak
to move one by himself; you must help him, Gwendolen."

Ralph put a wrap round Gwendolen, and, linking her arm in his, went
through the French window across the garden.

It was a glorious night. A full moon shed a mellow splendour across the
lawns, throwing the masses of the cedars into bold relief against the sky,
and glinting in all the diamond panes of the heavy-leaded windows. Over
the phloxes and tobacco plants that adorned the borders great moths were
wheeling, and bats were flickering in and out of the plantation that screened
the stables from the house. As the garden sloped towards the river the turf
was more closely shaven, and along the water's edge were sunk pots in
which magnificent geraniums and sweet heliotrope were growing.

Moored by the extreme boundary of the garden Ralph's house-boat lay;

it contained a little bedroom and two sitting-rooms, fragrant with flowers
and light with mirrors and thin curtains, and the upper part, covered in with
a pale green awning, was a mass of flowers and palms. Here were deck
chairs, and little tables, and Japanese lanterns.

Ralph put two chairs ready for Mrs. Austen and Sir Geoffrey, and then
looked at Gwendolen.
 "Shall we wait here for them, or would you like me to punt you up the
stream?"

"Let us stay here," she answered; "somehow——"

"Yes?" he said enquiringly.

"Somehow I fancy the others will not come," she said, rippling with
laughter. "Sir Geoffrey is always so thoughtful."

Ralph took her in his arms and kissed her passionately.

"I love you—I love you," he said, between set teeth, and Gwendolen
drew a sigh of perfect content. "If it could always be like this," he went on.
"Just you and I in the peace, with the river and the moonlight to reflect our
happiness."

But Gwendolen shook her head.

"You would soon tire of that," she said, and when he would have
demurred laid her hand upon his lips. "I hope you would, at any rate, for I
would not like you to be a lotos-eater dreaming your days away. There is so
much to do in the world, Ralph, and surely we, to whom so much has been
given, would not wish to give nothing in return."

He kissed the hand that caressed him.

"Tell me what I am to do."

Gwendolen considered.

"It is not easy to see just at first," she admitted, "but work, like charity,
begins at home. You will be a good master to your household, and will take
an active interest in the estate. You will be so anxious to make the tenants
happier in their respective stations that you will be surprised to find how
many things go to make up their lives. Life is a big bundle of little things,
you know, not a little bundle of big ones. If you really set your heart upon
doing good you will never stop for lack of something to do. That is a
wonderful thought, Ralph: there is no end to the good you can do in the
world."

"Go on," he said tenderly; "go on, dear, good little woman!"

"That is only thinking of your life at home," said Gwendolen; "but there
are wider interests outside. I should like you to make a name for yourself in
the great world; it might be in philanthropy, it might be in politics. I'm often
sorry you have no profession, but the world has always need of good men,
and I won't let you hold wool for me while the world wants one pair of
honest hands. Oh! Ralph, wouldn't that be more worth while than idling
your life away, even if it could always be like this?"

"Much more worth while," he answered gravely. "You have made me

happy; you will make me good; you may make me famous. That is a great
deal for one little woman to do for a man. What am I to do in return for
you?"

"Only love me," she said. "Love me always as you do now; never any
less tenderly or truly, even when the other interests are nearer than they are
to-night. What more can you do than give me love—the best thing in the
world?"

"I think I may safely promise that," Ralph said, and his deep voice
quivered. What had he done that Providence should heap blessings on him
so lavishly? For what had already been bestowed upon him he could never
show sufficient gratitude, and now there was the crowning gift of all—the
love of a pure and beautiful girl, whom he knew he had loved all his life.

Gwendolen lay back in one of the deck chairs, and Ralph, leaning
against the wooden railing, feasted his eyes upon the picture that she made.
In a dress of white mousseline-de-soie, trimmed with rare point lace, she
looked ethereally beautiful in this setting of coloured lamps and lovely
flowers. Her hands were clasped upon her lap, and the moonlight caught the
diamonds in the ring that he had given her, and even sought out the little
diamond drop that did duty as an earring. Against the scarlet cushions on
which she reclined her fair skin showed like ivory, and Ralph was filled
with something akin to amazement that this incarnation of all that was