120 E. A. MINClllN AND II. M. WOODCOCK. 



the nucleus, whereas isolated chroinatoid grains are a fainter 

 red. Figs. 16, 17 show instances of which we feel fairly 

 sure, as also figs. 4, 25 of a large form. In the latter the 

 two granules contrast distinctly with the mass of chromatoid 

 substance, of which we have more to say below. Their posi- 

 tion in this case, some distance from the nucleus, agrees 

 closely, it will be noticed, with that of the one grain in the 

 iron-ha3matoxyliu individual drawn in fig. 49. On the specially 

 thin smear, already alluded to, which was only lightly stained, 

 most of the parasites show no signs of these granules ; only 

 in one or two individuals is a round, faintly pink-staining 

 body present, which probably represents one (figs. 8, 9). 



There is another point to mention in this connection. In 

 Giemsa-stained smears, parasites which show these granules 

 clearly are much scarcer than in iron-ha.'matoxylin smears. The 

 explanation is probably as follows : To judge from the iron- 

 hajmatoxylin smears, in certain of the parasites the grains (or 

 grain) are more or less separated from the nucleus (cf . figs. 44, 

 46, 49), but in others they are close to it and may be in 

 contact with it. Hence in the Giemsa films, where the nucleus 

 is so obviously enlarged by being overloaded with stain, the 

 granules may be swamped, as it were, and not distinguishable 

 from the nuclear mass. We may expect, therefore, to be 

 able to recognise them only Avhere they are well out of the 

 nuclear zone. 



Since the granules appear distinctly smaller in Giemsa- 

 stained preparations than they do in iron-haematoxylinones, it 

 is most likely that the former depict more nearly the true size, 

 and that in the latter there is some overloading, due to the 

 strong affinity the granules have for the irou-hjematoxylin 

 stain. This is rather important in considering the significance 

 to be attached to these bodies. For if they were of chromatin 

 in the ordinary sense — like the grains and masses in the 

 nucleus — we should expect to find them apparently quite as 

 large in the Giemsa smears as in the iron-ha?matoxylin ones; 

 since, as is well known, the Romanowsky stain is deposited 

 most heavily in and around chromatic structures, with result- 



