368 REPORT— 1891. 



The best results I got by pouring in very slowly, a drop at a time, 

 about one every minute or so, 



96 per cent, methyl alcohol (CH3OH). . 10 vols. 

 Salt water ....... 90 vols. 



Natrium chloride . . . . 0-6 vol. 



After three-quarters of an hour or so, if they had not retracted, I 

 poured on quickly a large quantity of hot sublimate ; by this method I 

 succeeded in getting some specimens half retracted. 



For preserving specimens for sections I found the best results were 

 obtained from specimens treated for two minutes in 1 per cent, osmic 

 acid, then passed for two minutes through alcohol of 5, 10, 20, .30, 40 per 

 cent, up to 90 per cent., hardened in absolute alcohol, and imbedded in 

 paraffin. 1 found it best to leave the Spongicola in the sponge and dissect 

 it out after having hardened the whole in absolute alcohol. 



For staining I used mostly borax-carmine and hasmatoxalin, stain- 

 ing the SiMUfjicola whole after having dissected it from the sponge. I 

 did not obtain any good results from dissecting the Spongicola living, 

 as it seems capable of withdrawing itself to almost any extent inside 

 its tube, which in the interior of the sponge is very soft and easily 

 torn. 



History. — Spongicola fistularis was first discovered and named by 

 Professor All man in 1874 (Stephanoscyphus mirabilis, 'Nature,' July 

 30, 1874, ' Ann. and Mag. Nat. Hist.,' 4th series, vol. xiv., 1874, p. 237). 

 He describes it as inhabiting horny sponges in shallow water on the 

 south coast of France. 



From the fact that he could discover no hypostome or proboscis he 

 came to the conclusion it was not a true hydroid. He further made out 

 what he believed to be four longitudinal canals extending from the base 

 of the tentacle-crown some distance back and projecting into the interior 

 of the body-cavity. These, he says, are connected with a ' circular 

 canal' situated in the body-wall, ' which is wide, and easily admits a 

 needle.' This, he says, is continuous and without septa, having a distinct 

 endodermal lining. 



He failed to find an endodermal lining to the longitudinal canals, 

 though he thinks one probably exists. 



He further states that the tentacles are placed in ' two closely ap- 

 proximated and alternating series of 18 each, forming a single circlet,' 

 and that their structure is the same as that of a typical hydroid, that 

 when retracted the terminal orifice is closed over them, and that the 

 anterior part is thin-walled and very contractile, like a hydranth in its 

 hydro theca. 



He divides the animal into a proximal and distal portion, and says 

 that the mouth is probably situated where the two join, and that the 

 proximal cavity is the true digestive cavity, while the distal cavity is 

 homologous with the umbrella, and the tentacles with the marginal 

 tentacles of a medusa. 



He also states there is no endodermal lining to the distal cavity, 

 while the axial cavity has a well-marked one. 



Therefore, though the form and habit are those of a Hydroid Tropho- 

 some, its organisation is that of a medusa. So he says it is as far 

 removed from Hydi-oida as from Siphonophora, and he proposes therefore 



