752 MURIEL ROBEIJTSOX. 



obtains among" the group. Thus it was never found upon 

 Emyda (the milk tortoise) living in the same lake with the 

 Nicoria, nor upon thesiluroid fish Saccobranchus, nor upon 

 tlie water-snakes which shared the same habitat. Even in 

 an area so restricted as a well, these leeches were only found 

 to infest the Nicoria. Moreover, Ozobranchus lays its eggs 

 upon the carapace of the tortoise ; they ai'e of a dai-k brown 

 colour, closely resembling that of the tortoise, and are so 

 firmly cemented on that it requii-es a knife or some fairly 

 sharp instrument to detach them. It appears that the leeches 

 move readily enough from one tortoise to another, but it is 

 difficult to make out exactly how they are adapted to the 

 terrestrial night-wandering of their host. Tlie Nicoria spends 

 all the day sleeping in the water and comes to land to prowl 

 around at night, so most likely the leeches feed during the 

 day and drop off at night. Generally speaking I got more 

 leeches from nicoria caught in the evening, but there were, 

 however, some exceptions to this; presumably these were 

 cases where the tortoise had spent the night either in the 

 water or in a damp place. Ozobranchus is capable of 

 executing rather feeble swimming movements, and, in addi- 

 tion, can creep around upon its suckers in the usual way. The 

 time taken to digest a meal seems to vary from about three 

 to seven davs, according to the size of the leech. 



In Ozobranchus shipleyi the pi'oboscis leads into the 

 crop, which is a wide, very extensible sac dividing into two 

 large lobes at its lower end. The intestine opens from the crop 

 at the point where the division takes place. The upper end of 

 the intestine, which is rather wide, shows four long diverticula 

 on each side (see fig. in text, p. 753), This wide part of the 

 intestine terminates in a kind of chamber Avhich opens by a 

 narrow communication into a simple coiled tube, which leads 

 to the exterior at the anus. For some reason the most infected 

 part of the gut wall is almost always this chamber at the end 

 of the wide intestine. The accompanying diagram, which was 

 made from reconstructions of sections by the glass-plate 

 method, shows the relations of the vai-ious parts of the ali- 



