708 INTERNAL SECRETIONS AND PRESERVATION OF LIFE. 



to divide have generally the appearance of medium-sized hy- 

 aline cells, with a relatively large rounded nucleus and a com- 

 paratively small cell-body in which the mitoma is not easily 

 made out. But tKere is no doubt that cells with horseshoe- 

 shaped nuclei [the eosinophiles] divide, and that the nuclei 

 may even advance as far as the spirem stage without altering 

 their shape. Cells with more markedly polymorphous nuclei, 

 as, for instance, the ordinary oxyphile cells, certainly divide also, 

 but they seem generally to go through a preliminary rest- 

 ing stage in which the polymorphous nucleus returns to the 

 rounded form." 



In Gulland's plate, Figs. 3 and 6, which refer to eosinophiles 

 from newt's blood, graphically portray a secondary process 

 through which these cells can subdivide, or rather yield a por- 

 tion of their substance. In 3, a spherical pseudopod is in the 

 act of being formed; in 6, three similar masses appear, the 

 lowest of which is on the point of being separated by the 

 mother-cell. Referring to the bridges that connect net-works 

 of granules with basophile leucocytes, Gulland remarks: "I 

 have little doubt that when that stage is reached [he associates 

 the phenomenon with a supposed process of degeneration] these 

 bridges are torn across and the granules are actually left be- 

 hind. This forms an exact parallel to what happens in the 

 eosinophiles of the newt's blood." 



It thus becomes evident that recognized cytological phe- 

 nomena sustain the conclusion that neutrophile leucocytes are 

 the parent-cells of eosinophile leucocytes, and that eosinophiles can 

 part with segments of their cell-substance. 



But does the process of neutrophilic mitosis actually occur 

 in the liver? M. Duval, 36 in his study of the haematopoietic 

 functions of this organ, refers to the proportion of the red to 

 the white corpuscles in the blood of the portal vein as com- 

 pared to that in the hepatic vein, and writes: "Researches in 

 this connection give as result: 1 white corpuscle to 746 red in 

 the portal vein, and 1 white corpuscle to 170 red in the sub- 

 hepatic veins. This difference can only be due to a production 

 of white corpuscles in the liver or to a destruction of red cor- 



M. Duval: "Cours de Physiologic," p. 200. 



