260 



PAPERS FROM THE DEPARTMENT OF MARINE BIOLOGY. 



the total 0.6 per cent Si0 2 and (Al, Fe) 2 3 to the "clay." That leaves 0.9 

 per cent "clay" to account for. Then assuming a composition of 96 per 

 cent CaC0 3 and 4 per cent MgC0 3 the percentage of MgC0 3 becomes 0.046 

 or less than is being considered here. Comparing the calculated and directly 

 determined results then we have, for sample No. 27353*3: 



Following a similar procedure with sample No. 27337, line I, 200 feet 

 from shore, we have: 



J Ca organisms = Corals and mollusks ; Mg organisms = Alga; and foraminifera. 



Here there is even less system and therefore less basis for assumptions 

 as to the probable value of the CaC0 3 : MgC0 3 ratio in the silt and clay. 

 Foraminifera seemed abundant in the silt, but there is not in this sample, 

 as there was in No. 27353a, any indication of a tendency in the fine sand 

 towards their increase. Making the same assumptions as in No. 27353a, we 

 have (deducting 0.50 per cent out of the 0.63 per cent of Si0 2 and (Al, Fe) 2 3 ) 

 0.7 per cent "clay" to account for. Again, if we take CaC0 3 at 96 per cent 

 of the silt and the "clay," the residue referable to MgC0 3 becomes negligible. 

 The totals compared then are as follows, for sample No. 27337: 



Considering the results for both samples together, we see at once that in 

 No. 27337 there is a much closer agreement between calculated and directly 

 derived results than in No. 27353a. I do not know the explanation of this. 

 Undoubtedly by far the largest part of the discrepancy in both cases is due 

 to error in the assumed proportion of Corallinaceae to Halimeda — that is to 

 say, the Corallinaceae are usually more than half of the algae. I do not believe 

 that the greater experience brought to the study of No. 27337 is adequate to 

 account for an appreciable part of the greater agreement of the two results 

 in this portion. It is much more probable that the ratio of the two types of 

 algae to each other was more nearly the assumed ratio, that is, in this nearer 



