BLOOD-PARASITES OF FISHES OCCURIUNG AT ROVIGNO. 121 



ing enlargement of size (cf ., for instance^ the kinetonucleus 

 of a ti-ypanosome). The fact that this is obviously not the 

 case here suggests that the granules are not chromatic 

 structures. 



In this couuection certain of our Twort preparations are very 

 instructive. As has been shown by Minchin (I.e.) the Twort 

 stain is in one respect superior even to iron-haematoxylin, 

 in that chromatic elements can be distinguished from others 

 by the fact that they alone stain red, everything else being- 

 green. These particular Twort smears were examined soon 

 after being made to see if tliey showed granules corres- 

 ponding to the ones we had ah'eady found in iron-hsematoxylin 

 films. In several individuals one or two granules were seen, 

 which we regarded as those for which we sought. They 

 were small, however, and not particularly conspicuous. They 

 were very faintly stained red, nothing like so deeply or 

 sharply as the grains in the nucleus. Some of the individuals 

 showing these granules were noted aud sketched at the time, 

 and then the smears were put aside to work out on our return 

 to England. On finding the same parasites again recently, 

 in order to draw them, we could no longer see the red 

 granules in any of the individuals marked. All signs of 

 them have vanished, although the red of the nucleus has not 

 faded at all. In one or two cases, however, in about the 

 position which was occupied by the red granules (according 

 to our sketches), small rounded areas, somewhat diffuse in 

 outline, can be made out, staining a rather deeper green 

 than the surrounding cytoplasm (cf. fig. 59). It may be 

 that these greener areas mark the position of the structures 

 which stain so intensely with iron-hgematoxylin. In the great 

 majority of the individuals stained with Twort, however, 

 the body is uniformly pale green in colour, and cannot be 

 said to show any indications of the granules. 



To sum up, we regard the above-described characteristic 

 bodies as composed, at any rate, chiefly of achromatic material. 

 Our opinion is based on the one hand upon a comparison of 

 their staining reactions to iron-liEematoxylin and to Giemsa, and 



