Development of the Marine Sponges. 327 



end of the pipette, whose aperture should be large enough to 

 admit the embryo, must be put into the water close to it, while 

 the finger is pressed tightly upon the opposite end ; then, when 

 the finger is withdrawn, the embryos will, by the capillary 

 attraction of the water, pass up into the mouth of the tube, 

 and, by subsequently gravitating towards the lower part, 

 when the pipette is held upright, may thus be transferred to 

 a slide for examination or to a vessel for holding them with- 

 out any further force. They will live for several days in pure 

 sea-water, changed now and then for a fresh supply ; and 

 their whole development into the perfect sponge may thus be 

 easily followed ; but, as this process is not stationary, the 

 embryo itself can only be expected to be in its perfect condition 

 at the moment it issues from the parent sponge, and must then 

 be examined for this purpose, as it changes somewhat every 

 hour afterwards. 



It has been above stated that to give the whole of the de- 

 velopment of the ovule, from its first appearance to the fully 

 developed sponge, I have had to study it in two species ; that 

 is, that, owing to the transparency of the ovules in the non- 

 spiculous sponge Halisarca lobidaris ) the segmentation of the 

 yelk is much better seen here in situ than in Halichondria 

 simulans, where the ovum is opaque ; while, as regards the 

 embryonal form and subsequently developed sponge, not only 

 its larger size, but the presence of spicules in the embryo, and 

 their final arrangement into the skeleton-structure of Hali- 

 chondria simidans, render the embryo of this sponge much 

 more eligible for this part of the development than that of 

 Halisarca lobidaris, in which there are no such aids. 



We shall first, then, commence with the development of the 

 ovule from its earliest appearance to the ultimate segmentation 

 of the yelk in Halisarca lobidaris, and then follow the deve- 

 lopment of the embryo into the perfect sponge in Halichondria 

 simulans, comparing the latter afterwards with the develop- 

 ment of the embryo of Grantia compressa into its perfected 

 form, thus supplying, to a certain extent, that detail of which 

 the late Prof. B,. E. Grant discovered the salient points forty- 

 seven years since (' Edinb. Phil. Journ.'), and taught them to 

 me, his friend and pupil, in his lectures thirty-three years ago 

 at University College. 



It should be here mentioned that this subject is not entirely 

 new to me, as I have already described and figured the ovum, in 

 Tethya cranium and T. zetlandica respectively, from preserved 

 specimens, in which it had probably advanced to very nearly 

 the full period of embryonal development (' Annals,' 1872, 

 vol. ix. p. 409, pi. xxii.). 



