APPENDIX II 415 



a day's difference having been observed; and we may infer that 

 these diseases, and probably others also, cause a loss of vitality in 

 the patient's leucocytes, so that by this procedure the loss of vitality 

 can be measured. It is important to remember this point, for if 

 the making of a measurement is delayed it may be found that all 

 the cells are dead in both control and test preparations. This 

 method of measuring the lives of leucocytes may also prove of 

 value in prognosis as well as in diagnosis. 



I do not think that any difficulty will be met with in making 

 the counts, with the exception of a possible one caused by the 

 agglutination of the leucocytes. Occasionally large clumps are met 

 with. If the cells are clumped, however, it does not necessarily 

 follow that they are dead far from it, for they may be very active, 

 though I am of opinion that if clumped death will soon occur. 

 The cells in a clump can generally be counted. Ruptured cells 

 are counted as dead. If bacteria are seen in large numbers in a 

 film the capillary tube is discarded. The revolving apparatus is 

 not essential, but more constant results have been obtained by its 

 use. As far as possible I have purposely avoided handling the 

 blood of the person whose leucocytes are to be tested, for fear of 

 injuring the cells. The variations of the alkalinity of the plasma 

 may, I think, be neglected, as it is not sufficient materially to alter 

 the length of the lives of the cells. This is borne out by the 

 experiments with cancer plasma, because that plasma is more alka- 

 line than normal and yet does not shorten life. 



Summary 



I fear that it is too early to arrive at any definite conclusions 

 from so small a number of experiments, but I think that there 

 publication is justified in order to explain the method employed 

 and because the results are sufficiently promising to warrant further 

 investigation, though the work must still be regarded as being in 

 the experimental stage. I hope that this method will be tried by 

 others, as the problem given in the enunciation may lead to im- 

 portant developments, and especially as this kind of research in- 

 volves the striking of averages and a large amount of experiment to 



