CONTRIBUTION TO AN EVALUATION OF THE SEA. 315 



embryos produced in a year, and from the fact that specimens only 

 1-5 cms. long have been found to contain fully developed free embryos 

 in the inner flagellated chambers in the aul^umn I think it very probable 

 that this species might in favourable seasons yield three generations 

 within a year, rushing through two of them in late autumn. 



Similar breeding phenomena to those described in S. coronatum 

 are presented by G. comj^ressa and Leucosolenia. Tiny specimens 

 of G. compressa 1-3 cms. long taken in October at an age not greater 

 than 7 months and of an actual age of less than 4 or 5 months have been 

 found to be crammed full of embryos, while large specimens 8 cms. long 

 with an average width of 3-5 cms. taken in March and April have been 

 found to contain only immature ova. In the following June, however, 

 specimens similar to the latter have yielded crowds of embryos. Cor- 

 responding results to these have been obtained with Leucosolenia. 

 Thus the summer and winter breeding periods appear to be general in 

 our calcareous sponges. The life-cycle of Grantia has been carefully 

 followed round the year, and it was observed that in one situation where 

 this sponge was extremely common of a large size in June they gradually 

 disintegrated during July, so that by August it was impossible to find 

 any but the tiniest specimens, which were doubtless the first comers of 

 the new summer crop. The same difficulty in finding any but small 

 specimens was also met with on other grounds in the district w^here 

 these sponges can generally be obtained at any other time of the 

 year. 



The summer crop continues to grow during the autumn, and by about 

 December may attain to an area in side view, i.e. on one flat side, of 

 about 10 sq. cms. By the following March specimens may be found 

 with an area of from 25 to 30 sq. cms.,* and at the end of a year's growth 

 in the following June specimens of average size may attain to an area 

 of at least 30 sq. cms.* in side view, or ni other words, the whole " body- 

 wall " would have an area of more than 60 sq. cms. After delivering 

 their summer embryos these specimens, as we have already seen, die 

 down and a fresh crop arises. 



Of siliceous sponges few records of much value have so far been 

 obtained, but a colony of Halicliondria fanicea grew on a flat surface 

 between June and the following January to an area of about 45 sq. cms. 

 and of about the usual thickness of this sponge. 



* More exact measurements will be given later. 



