284 THE DIVISION OF LEUCOCYTES 



22 of whom were suffering from undoubted 1 cancer, 2 

 and 47 of whom apparently were not. Of the latter, 

 which for convenience will be called the "control" cases, 

 some were "healthy" and others were suffering from 

 various diseases (hospital patients) . We did not count 

 the granules in each of the 235 negatives as the latter 

 were obtained, but the plates were developed and then 

 put away until a hundred or more had collected. The 

 name of the person from whom the blood had been 

 taken w r as entered into a book with the age, sex, 

 disease (if any), and other details. The negatives were 

 numbered consecutively, and the numbers corresponded 

 with similar ones in the book against the names of the 

 persons from whom the cells had been derived. The 

 samples of blood were taken from persons, cancerous or 

 otherwise, as they came into hospital, and therefore, 

 without referring to the book, a number on a negative 

 gave no indication from whom the cell it depicted was 

 derived. With three exceptions, the samples of blood 

 were collected and photographed by one of us, who 

 kept the book in his laboratory. The counting w r as 

 done by another, who had no idea to whom the 

 numbers on the negatives referred. 



The only possible source of error is in the counting. 

 Some of the negatives were not quite perfect, and 

 ' some of the granules appeared blurred ; hence there may 

 be a small error in some of the numbers, but it cannot 

 be very important judging by the uniformity of the 

 averages. 



1 Determined either by such clinical manifestations as recurrence or 

 metastasis, or by pathological examinations. 2 Carcinoma. 



