258 



solution with great success. Of the several modifications of the Ro- 

 mano wski method I have combined those of Lave ran [8] and 

 Plimmer [11], namely, using Bleu Borrel, erythrosin and tannin orange, 

 after fixation with absolute alcohol. The use of the tannin orange is op- 

 tional. This method gave good results for the detection of the parasite 

 and its general structure, though for the nuclear (chromatic) detail of the 

 parasite absolute alcohol is perhaps not an ideal fixative. I have found 

 a mixture of saturated aqueous corrosive sublimate two parts, with 

 absolute alcohol one part, as suggested by Hin tz e [4] after Schau - 

 dinn , to give very good results as a fixative. The preparations may then 

 be stained with Delafield's haematoxylin , Heidenhain's iron- 

 haematoxylin, or gentian-violet, using eosin as a successful plasma stain, 



Fiff. 1. 



Fig. 2. 



Fig. 1. Blood coi'puscle of newt, with vei'miform haemogregarine in situ; 



X 2100 approx. 

 Fig. 2. Same, with schizont dividing into merozoites; x 2100 approx. 



I have also tried the Romanow ski stain after corrosive sublimate and 

 alcohol fixation with fair results. It may be here noted that the haema- 

 toxylins already mentioned require a long time for action in order to 

 stain the parasite at all deeply. It is not too long to leave the films for 

 several hours in concentrated Delafield's haematoxylin, while from 

 sixteen to twenty four hours is necessary with a dilute acidulated solu- 

 tion of this stain, even longer than advocated by Hintze in the case of 

 L. ranarum. About the same time is required for the action of these 

 stains after fixation with osmic acid, or mixtures containing it. 



Some time after beginning this work my attention was dra-^Ti to 



