604 W. J. SOLLAS. 



includingPachymatisma Johustoni^ Tethya lyncuriurU; 

 a massive orange-coloured Suberites^ various Calcispongise, 

 and Halisarca Dujardini and lobularis. 



Judging from the appearance of the grotto it would be 

 difficult to find a more " healthy " locality for sponge growth. 



The specimens when gathered were immediately placed in 

 clear sea-water and transferred to fresh sea-water in clean 

 " cuvettes " on reaching the station. They were not allowed to 

 remain there for many hours, but as soon as convenient were 

 placed in a weak solution of chromic acid, to which a few drops 

 of osmic acid had been added (some of the specimens from the 

 rocks of Roscoff were placed in a solution of mercuric chloride 

 and a little acetic acid). They were finally brought into 

 absolute alcohol, in which medium they were kept till they 

 were cut and mounted in England. The cutting was accom- 

 plished both by the freezing and paraffin methods. In the 

 latter case the specimens were embedded sometimes by the oil 

 of cloves, and sometimes by the chloroform (Griesbach) pro- 

 cess. The sections were stained by immersion first in eosin, 

 and subsequently in hsematoxylin ; the best results were ob- 

 tained when the staining was postponed till after cutting, when 

 the sections were arranged in series and fixed to the glass slide 

 by caoutchouc (Threlfall's method). Notwithstanding the 

 various kinds of treatment to which they have been subjected, 

 all the slices show essentially the same characters. The only 

 discernible point of diff'erence lying in the fact that slices 

 obtained by freezing have not suffered any contraction, while 

 those cut in paraffin have shrunk, notwithstanding all precau- 

 tions, very appreciably. Since the shrinking has evidently, 

 however, been uniform, it does not affect the characters of the 

 embryos except as regards their apparent size. As the sections 

 represented in the plate have all been drawn by camera lucida 

 from paraffin-cut specimens, they are slightly smaller than 

 they should be. 



I am particular in entering into these details in order to 

 show that the appearances which I have next to describe are 

 not easily explicable, either as pathological or artificially pro- 



