GANYMEDES ANASPIDIS (noV. GEN., NOV. SP.). 159 



Finally, a certain number of the Gregarines are found 

 associated in pairs, the attachment being by dissimilar ends 

 (text-fig. 1). Not very many are in this state, but I suspect 

 that the shock of killing, and the subsequent manipulation, 

 manage to sever the connection between a large number of 

 couples, and therefore cannot say if association always super- 

 venes when tlie parasites reach a certain size, nor what are 

 the proportion of couples to free Gregarines. 



Association marks the close of tlie first tropliic period. 

 In the second trophic period the Gregarines are non- 

 motile, have lost all the complex structure they had before, 

 and are characterised by their (probably rapid) growth to a 

 very large size. In this state they are found in the liver- 

 tubes, of which there are twenty or thirty, lying free in the 

 hasmococoele, and not intertwined. It follows that the asso- 

 ciated couples must migrate forwards to the junction of mid- 

 and fore-gut, where the liver tubes open, and thence back 

 into one of these. On penetrating a safe distance along the 

 tube, a transformation must take place, the two Gregarines 

 undergoing complete cytoplasmic fusion, a state of affairs 

 known hitherto only in those neogamous forms from Holo- 

 thurians, Cystobia and allied genera (Woodcock, 6). 



These fused couples, looking just like one cell with two 

 nuclei, are found wedged in between the cells of the wall, 

 with a considerable free surface for absorption towards the 

 lumen of the tube. There is often another free sui'face on 

 the exterior, due, I should say, simply to the growth of the 

 creature, and the consequent forcing apart of the liver-cells 

 (text-fig. 3). For this growth, Ganymedes is here in a very 

 favourable place, since the so-called liver, in addition to 

 producing digestive fei'ments, is the organ where a great 

 part of the food is absorbed ; and so, while the parasites may 

 enter on this period when measuring no more than 70 x 60 fx, 

 they often attain to the considerable size of 200 x 130 fi, and 

 I have seen one that measured 300 jn in its greatest length, 

 though its breadth was only 120 ^u. The shape is variable, 

 from a nearly perfect sphere to a long ellipsoid or ovoid. 



