190 



H. S. H. WARDLAW. 



are grouped about the most probable values are also well 

 shown by the different heights of the maxima on the curves, 

 and the different degree of steepness with which they rise 

 to these maxima. 



The striking difference between the distribution of the 

 percentages of fat and those of the percentages of the other 

 constituents is not peculiar to the present results, but is 

 equally well shown by the results of Leeds, Carter and 

 Droop Richmond, and Oamerer and Soldner already quoted 

 in this paper. The following are the relative frequencies 

 of occurrence of the various percentages of fat recorded in 

 their papers arranged as described above. In these cases, 

 of course, since the total numbers of results are not one 

 hundred, the numbers given do not represent percentages. 



Per- 































cent. 



i-h 



and 



1-8 



2-1 



2-4 



2V 



3-0 



3-3 



3-6 



3-9 



4-2 



4-5 



4-8 



5-1 



c>-4 



and 



of fat. 



under 



























over 



Num- 



*1 



3 



2 



5 



6 



6 



9 



7 



2 



5 



3 







2 



1 



ber of 



to 







3 



3 



6 



6 



2 



7 



5 



5 



3 



4 







13 



results 



tie 



6 



6 



10 



6 



11 



9 



9 



7 



3 



3 



2 



1 



4 



* Camerer and Soldner. f Leeds, J Carter and Droop Kichmond. 



As will be seen from these figures, the results show 

 practically no tendency to group themselves about any 

 particular value but occur with fairly even frequency over 

 a wide range of values. 



When, however, the percentages of the other constituents 

 are arranged in the same way they are found, as in the case 

 of the results of the present work, to group themselves 

 about certain values with varying degrees of closeness. 



The process by which the fat of human milk is secreted, 

 therefore, unlike the processes by which the other con- 

 stituents are produced, does not seem to favour the appear- 

 ance of milk containing any particular percentage more 

 than any other between certain wide limits. Although it 

 might be contended that the irregularities found in the 



