Spermatozoa in the Spongida. 109 



towards the centre or at the circumference I could not deter- 

 mine, projected a single cilium. The body while in activity 

 exhibited a subpolymorphic form and often became cup-shaped 

 or conical towards the middle on one side (fig. 26, g) ; while 

 the rim as often became thicker and seemed to hold the granule 

 just opposite the cilium, giving the translucent centre a kind 

 of horseshoe-shape (fig. 26, e). 



The cilium propelled the body forwards, but as often too 

 presented a bulbous soft swelling at the end, which seemed to 

 act as a sucker in anchoring the body to other cells and sar- 

 codal objects in the field (fig. 26, h). After a time, when the 

 body was still, the whole became indolently polymorphic 

 and amoeboid in shape, while the cilium shrunk up to a short 

 process (fig. 26, i). They were often seen in twos and fours 

 together in a flexible mother cell about the field of observation 

 (fig. 25) ; and each individual measured l-6000th of an inch 

 for the diameter of the body and 5-6000ths of an inch for the 

 cilium. 



These bodies were numerous while the Halisarca was quite 

 fresh — that is, on the first day of capture and for three or four 

 days afterwards, when it disappeared, as vibrios and other 

 monadine bodies announced coming decadence in the sponge. 

 While their appearance in this sponge in its fresh state, 

 together with the presence of the ova considerably advanced, 

 led me at first to think they might be the spermatozoa of the 

 species, the bulbous inflation of the tail and its power of 

 anchoring the body looked so monadine that, together with its 

 unusual appearance, whether spermatozoon or monad, its publi- 

 cation may not be altogether useless. Having since seen the 

 free ends of the ectodermal cells of a Gastrula become bulbous, 

 this alone is only proof of their sarcodal, polymorphic com- 

 position. 



I have stated that the specimen of Halisarca, Dujardinii in 

 which these bodies were exclusively observed was " as pregnant 

 with ova as that of Orantia compressa, but not so far advanced;" 

 that is to say, the ova were in that stage of development when 

 the granular yelk is clearly seen investing the nucleus, nu- 

 cleolus, and germinal vesicle, measuring about 7-6000ths of 

 an inch in diameter, and still reptant ; while in a specimen 

 which came from the Isle of Man and was sent to me by Mr. 

 T. Higgin of Liverpool, they measure 45-6000ths of an inch 

 in diameter (nearly seven times as much), and thus are visible 

 to the naked eye, having passed through the nuclear stage 

 and become probably nearly ready for delivery. 



Hence, as the largest specimens of Gastrula in Grantia com- 

 pressa, which appear to be equally ready for exit, do not 



