78 



MARINE BOTTOM SAMPLES CF LAST CRUISE OF CARNEGIE 



Table 26. CaCOs content of different size grades in certain calcareous samples 



Sample 

 no. 



Size grade and CaCOs per cent of grade (A) and of sample (B) 



.25 - .12 mm 



.12 - .07 mm 



.03 mm 



.03 - .005 mm 



< .005 mm 



♦Although the method used is accurate only for comparison purposes, the sum of the CaCOs found, 

 plus that estimated microscopically for the remaining sand grades, agrees well with the value of total 

 CaC03 given in table 1, for all samples e.xcept no. 17. The discrepancy in this sample is owing to the 

 fact that the part used in chemical analysis did not contain as great amounts of volcanic glass frag- 

 ments as did the part which was mechanically analyzed. 



carbonate content in the different size grades of certain 

 Globigerina sediments were carried out by a titration 

 method similar to that described by Trask (1932). The 

 results indicate the presence in some samples of double 

 maxima in the proportions of calcium carbonate in the 

 different size grades. This is shown particularly well in 

 sample 17, a volcanic ferruginous ooze from the south- 

 east Pacific. The sand grades of this sample contain ap- 

 proximately 95 per cent calcium carbonate and the clay 

 grades about 75 per cent. The silt grades, on the other 

 hand, contain only 58 per cent CaC03. The top part of 

 sample 20 similarly shows less CaCOs in the silt grade 

 than in the clay, and CaCOs increases in a regular man- 

 ner from the sand to the clay in sample 53. In the re- 

 maining samples--20 (bottom), 24, 40, and 82--the cal- 

 cium carbonate, as is to be expected, decreases with 

 decreasing particle size. If all the calcium carbonate in 

 these sediments comes from the breaking up of the 

 shells of forarainifera, it is difficult to understand the 

 size distribution of CaC03 in samples 17, 20, and 53. A 

 possible explanation for this distribution is that the fine- 

 grained calcium carbonate has a different origin from 

 that of the sand grades. 



In general, the median particle diameters of deep- 

 sea sediments vary directly with the amounts of calcium 

 carbonate in them. This relation is well shown by the 

 samples from the south Pacific which were analyzed by 

 the pipette method, as may be seen from figure 28 in 

 which the logarithms of the median diameters of sam- 

 ples 19, 21, 31, 34, 40, 43, and 81 are plotted against the 

 respective contents of calcium carbonate. Correns has 

 pointed out a similar relation for the deep-sea sedi- 

 ments of the south Atlantic. 



Bureau of Soils' Method 



Besides the precise analyses by the pipette method 

 described above, forty-one mechanical analyses by the 

 older method of the United States Bureau of Soils were 

 carried out by A. H. Barker and the author. The results 

 as percentages in various size grades are given in table 

 27 and are plotted in figures 31 to 36. The method used 

 is quite familiar and it need only be pointed out that the 

 errors involved are of a far greater order of magnitude 

 than those of the pipette method. For coarse-grained 

 samples, however, the method is fairly satisfactory. 



In table 28 the magnitude of the errors due to sieving 

 and quartering in the mechanical analyses of the sand 

 grades of two Globigerina oozes, nos. 19 and 21, is 

 shown. The probable error due to sieving was found to 

 vary between 0.2 and 0.5 per cent; that due to quartering 

 is close to 0.4 per cent. The probable error of a single 

 determination by sieving of the percentages in any sand 

 size grade thus varies between 0.4 and 0.6 per cent. 

 That these differences are of little significance in the 

 interpretation of the analyses may be seen from figures 

 29 and 30, in which histograms are shown of the most 

 widely differing of the multiple analyses of the sand 

 grades of these two samples. Compare, in this connec- 

 tion, the results obtained by Wentworth (1927). 



In figure 31 are shown the histograms from mechan- 

 ical analyses, both by the Bureau of Soils' method and 

 the pipette method, of thirteen Globigerina oozes from 

 the south Pacific which show double maxima in the size 

 distribution, one in the sand and the other in the clay 

 grade. It has been pointed out already that this size 

 distribution may reflect a similar distribution of the 

 calcium carbonate content. 



The remaining Globigerina oozes, for which histo- 

 grams are shown in figure 32, may be divided into four 

 groups as follows: (1) a group of six samples of high 

 CaCOs content in which there is a single maximum in 

 the sand grades owing to the preponderance of unbroken 

 shells of pelagic foraminifera; (2) a group of four 

 samples of intermediate calcium carbonate content in 

 which there is a single maximum in the fine grades ow- 

 ing to the presence of much finely divided calcium car- 

 bonate as well as of fine-grained noncalcareous material; 

 (3) three sainples--nos. 15 and 17, Globigerina oozes 

 from the south Pacific, and no. 53, a volcanic Globiger- 

 ina mud from the northwest Pacific--which contain a 

 relatively large amount of volcanic debris of gravel or 

 sand size in addition to pelagic foraminifera and fine 

 material, hence are poorly sorted; (4) a group of three 

 samples which were not satisfactorily dispersed because 

 of the presence of large amounts of iron and manganese 

 oxide. These are included simply to indicate the appear- 

 ance of such unsatisfactory mechanical analyses of Glo- 

 bigerina oozes. 



Histograms for all the samples of north Pacific clay 

 which were mechanically analyzed either by the pipette 

 or the Bureau of Soils' methods are shown in figure 33. 

 The difficulty of interpretation of the analyses made by 



