24 Art. I- — N. Annandale and T. Kawainura : 



In /S y . dementis, remarkable in this respect as in so many- 

 others, the production of gemmules does not appear to be a seasonal 

 event, but to be brought about by the physiological changes, as to 

 the nature and cause of which we have as yet no precise informa- 

 tion. All that we can say is, that out of many hundreds of speci- 

 mens of this species, only two which were taken in a single haul 

 of the net in the first week of October contained gemmules ; that 

 both were attached to living shells of Corbicula, and that in both 

 of them the sponge was completely degenerate. 



As to sexual reproduction in the Spongillidae of L. Biwa we 

 have little precise information, except that larvae are produced in 

 great abundance by $. dementis in the latter part of September and 

 the first half of October, and more sparingly at the beginning of 

 November. It is probable that this is the case also in other species 

 as there appears to be a somewhat luxuriant new growth, in parti- 

 cular of S. semispongilla, at this time of year. 



The bionomical relations between fresh -water sponges and 

 other organisms are still for the most part imperfectly understood. 

 In L. Biwa a small Oligochaet worm of the genus Chaetogaster is 

 not uncommon in the canals of S. semispongilla growing on weeds 

 in the South Lake. Apparently it avoids S. fragilis in close proxi- 

 mity and, unlike Gh. spongillae Axnandale 1} (which lives only in 

 dead or dying sponges), is found in specimens in full vegetative 

 vigour, even before the appearance of gemmules. S. Matsumura 

 has described an interesting Heteropteron (Aphelocheirus kawamurae 

 Matsum.) 2) as yet found only in a cavity on the surface of the semi- 

 massive phase of S. dementis collected at Seta, while we noticed a 

 suctorial Protozoon {Sphaerophrya sp.) abundant on the dermal 

 membrane and a small Gammarid in the canals of the massive form 

 of the same species gathered off Komatsu in November, 1915. This 

 phase of the species, although it sometimes grows on small stones, 

 is so frequently attached to the shells of molluscs of the genera 

 Corbicula, Nodular ia, Melania and Vivipara that it is known to the 

 fisherman at several places as Kai-no-lcuso, or molluscs' excreta. 



1) Journ. As. Soc. Bengal, (n.s.) II, p. 189 (1906). 



2) Entomol. Mag. Japan, Vol. I, No. 3, p. 104 (1915). 



