172 H. S. H. WARDLAW. 



age and of the number of pregnancies of the mother, and 

 of the time of lactation, on the composition of the milk has 

 been studied. The effect of the length of the period of rest 

 which has elapsed since the last withdrawal of milk from 

 the breast and of the volume of the sample, and the differ- 

 ence between the average compositions of the milk from 

 each breast, have also been examined. 



The most striking feature of the results obtained is that 

 they seem to give further support to the view that the 

 secretion of the fat of human milk occurs independently of 

 that of the other constituents, since the percentages of the 

 latter show marked tendencies to approach certain "most 

 probable" values, the percentages of fat show not such 

 tendency, but the occurrence of any one percentage within 

 wide limits seems to be no more probable than that of any 

 other. 



2. The samples of human milk and methods of analysis. 

 The samples of human milk of which the analyses are 

 given in this paper were obtained from patients of the 

 Royal Alexandra Hospital for Women, Sydney, through the 

 courtesy of Miss E. M. Buckley, m.b., Pathologist to the 

 Hospital, to whom I express my best thanks. 



The patients who come to this hospital for confinement 

 remain there in the ordinary course of events for ten or 

 eleven days after the birth of the child, and of the 105 

 samples of human milk of which the analyses are given 

 only one represents the milk secreted after a period longer 

 than this after child-birth. The great majority of the 

 samples were obtained four or five days after the birth of 

 the child. 



The nurses who obtained the milk were at first instructed 

 to obtain it by massaging the breast of the patient, as a 

 better sample is said to be obtained in this way than with 

 a pump (Engel), but as the nurses seemed to have great 



