SUCTORIA l6l 



the cell-body. The mechanism of suction is doubtful ; but 

 from the way particles from a little distance flow into the open 

 funnels of Choanophrya, it may be the result of an increase of 

 osmotic pressure. The external pellicle of the tentacles is 

 marked by a spiral constriction/ which may be prolonged over 

 the part included in the sheath. The endosarc is rich in oil- 

 drops, often coloured, and in proteid granules which sometimes 

 absorb stains so readily as to have been named " tinctin bodies." 

 It usually contains at least one contractile vacuole. 



In Dendrocometes (and perhaps others) the whole cell may 

 become ciliated, detach itself and swim off; this it does when 

 its host (Gammarus) moults its cuticle. 



In fission or budding we have to distinguish many modes. 

 (1) In the simplest, after the nuclear apparatus has divided, the 

 cell divides transversely ; the distal half acquires cilia and swims 

 off to attach itself elsewhere, while the proximal remains 

 attached. The tentacles have previously disappeared and have 

 to be formed afresh in both. (2) More commonly fission passes 

 into budding on the distal face ; a sort of groove deepens around 

 a central prominence which becomes the ciliated larva (Fig. 62, 

 em) ; the tentacles of the " parent " are retained. This process 

 passes into (3) " internal budding," where a minute pit leads 

 into a bottle-shaped cavity.^ (4) Again, the budding may be 

 multiple, the meganucleus protruding a branch for each bud, 

 while the micronucleus, by successive divisions, affords the supply 

 requisite. Sphaerophrya (Fig. 61, e) and Undosphaera multiply 

 freely by fission within their Ciliate hosts, and were indeed 

 described by Stein as stages in their life-cycle. Conjugation of 

 the same type as in most Ciliates has been fully worked out in 

 Dendrocometes alone, by Hickson,^ who has found the meganuclei 

 (though destined to disorganisation) conjugate for a short time 

 by the bridge of communication before the reciprocal conjugation 

 of the micronuclei. 



We have referred to the endoparasitism of two genera. 

 Amoehophrya lives in several Acanthometrids, and in the aberrant 

 Eadiolarian Sticholonche (see p. 86). The attached species are 



' The spiral ridge figured by Hertwig (Fig. 61, i, o) is probably an incorrect re- 

 presentation of this structure, exceedingly minute in all genera but Choanophrya. 



2 In Choanophrya I have failed to find any pore, and believe the bud-formation 

 to be strictly endogenous. 



= See Quart. Journ. Micr. Sc. xlv. 1902, p. 325. 



VOL. I M 



