534 Mr. T. Bentham on some 



quite resistant to the influence of the haemolysin, and fall 

 distinctly into two categories. Short forms are present of 

 the usual Hsemogregarine type, namely, bean-shaped and 

 rounded at both ends, with a more or less centrally placed 

 nucleus of the diffuse type. They measure roughly 10 fi in 

 length by 3 p in breadth, and inter se do not vary very 

 much in size. These forms were taken to be the micro- 

 gametocytes, or those destined to give rise to the male forms 

 of the parasite. They were further characterized by being- 

 free from all metachromatinic or volutin granules, and in this 

 character they differ from the intra- and extra-cellular para- 

 sites found in the peripheral blood of Cottus (see PI. XVI. 

 fig. 1, 3b). The female elements or macrogametocytes were 

 elongate vermicular forms — as a rule, about 15 \x by 2 to 3 /j, 

 in measurement, and having a compact vesicular nucleus 

 situated about the middle of the cell-body. This nucleus 

 usually contained a large distinct karyosome situate within 

 the organ. No external chromatin grains were to be seen, 

 and, as in the microgametocytes, the cytoplasm was free from 

 volutin. The rest of the life-history in the leech could not 

 be made out with certainty, but there were present in the gut 

 a fair number of large rounded bodies having a more or less 

 compact nucleus. In some cases what appeared to be an 

 extrusion of karyosomes could be made out. These large 

 forms were taken to be macrogametocytes which had passed 

 into the macrogamete stage. No stages of the formation of 

 microgametes could be seen (see PI. XVI. fig. 1, 2 a & 2b). 

 It was at first thought that the elongate vermicular forms 

 might possibly be Ciithidial stages of Trypanosomes present 

 in the leech-gut, and to this end smears were stained with 

 iron hematoxylin. With this process the structure of the 

 nucleus was seen to approximate more closely to the usual 

 Hsemogregarine type, rather than to the appearance of the 

 same structure in a Trypanosome. From these observations 

 it is therefore fairly evident that schizogony takes place, not 

 in the blood of Cottus, but in the spleen, the gamogonous 

 cycle taking place in the gut of the invertebrate host. It is 

 also evident that the spores of the genus Hcemogregarina are 

 very resistant of death and dissolution in the invertebrate 

 host, since they seem to persist for a very long time in the 

 dorsal blood-vessel of the leech. This is substantiated by 

 the following facts. A cat-fish (Anarrhichas lupus) was 

 caught on one of our travvling-trips near Holy Island on 

 June 25th, 1913. Several specimens of the leech Ichthyo- 

 bdella anarrhichce were taken from the gills and placed in 

 a glass vessel in the aquarium. These leeches were kept 



