H. A. Baker — Loose Arenaceous Sediments. 831 



when the sample is in the tube and the remainder of the tube is filled 

 with water up to the level to which the bottom of the stopper 

 reaches. The weighing can conveniently be obtained by means of 

 a large specific gravity balance having one of the pans at a higher 

 level than the other and equipped with a hook underneath. After 

 the elutriation another weighing of tube and contents is taken, and 

 the next stage in the analysis can be proceeded with immediately, 

 whilst the weight of sediment remaining in the tube at the end of the 

 previous stage is being calculated. This required weight is calculated 

 from a formula which is obtained as follows : — 



Before jLet weight of sediment in tube = a gms. 

 elutriation (Let weight of water in tube =6 ,, 



After (Let weight of sediment in tube = c ,, 

 elutriation I Let weight of water in tube = r? ,, 



a is known when put in, and total weight of contents is found, 

 hence h = total weight contents — a. 



The total weight c -{- dis found by weighing. 



Put c -\- cl =^ x; then cl =^ x — c. 



Now the total volume of sediment and water remains the same at 

 the end of the experiment as it was at the beginning. 



Put p\ = density of water used, and 

 p^ = density of sediment. 



Take po, — 2" 65 (treating the sediment as a pure quartz sand). 



Then volume before elutriation = ^r-^ -\ cu. cm. 



2-bD pi 



c d 



and volume after elutriation = x-^^ -\ — cu. cm. 



2-65 



a h c . d 



%6E /^ " 2^65 pi 



Pi 



i.e. api + 2-656 — cpi + 2-65f/, 



and substituting x — c for d, we get 



ap, + 2-656 = cpi + 2-65a; — 2-65c. 



Hencec^ ^-^^i^-^'^-"^^- 

 2-65 — />i 



The formula gives very good results, since a, x, and 6 are all 

 determined by weighing, and the error involved by taking the density 

 of the sediment as 2"65 is very small, since arenaceous sediments are 

 very largely quartzose. For the value of pi appropriate to the 

 temperature of the water at the time of the experiment, it is con- 

 venient to consult a graph showing the variation of the density of 

 the water in use with change of temperature. This graph can be 

 constructed from the results of preliminary experiments on the water. 

 It may be remarked, however, that an error in the determination of 

 Pi considerably greater than the experimenter is likely to make in 

 practice, introduces only an insignificant error into the final 

 evaluation of c. It is well, however, to be as accurate as possible, 



