456 SUMMARY OF CUKEENT RESEAECHES RELATING TO 



it is suggested that the collared larvae in question were only separated 

 portions of a collared chamber layer. The development of these 

 larvBB, if such they are, is not, according to Saville Kent, the result 

 of segmentation, but is parallel to the colony formation observed in 

 various Flagellata, and especially in the newly discovered Choanoflagel- 

 late Salpingoeca fusiformis, in which a typical individual retracts its 

 collar and flagellum, becomes amoeboid, passes into a spherical quiescent 

 stage, and undergoes regular division ending in ciliated swarm-spores 

 which leave the capsule and develope into Salpingoecas. In the sponge 

 a parallel process results in ciliated and then collared individuals 

 which remain, however, united by a common gelatinous supporting 

 substance, which indeed occurs in the new Choanoflagellate Proto- 

 spongia hseckelii : a slight modification in the disposition of the zooids 

 in the latter would produce a very simple sponge. Prof. Schulze 

 emphasizes in answer, inter alia, the now well-established character 

 of the ground-tissue in sponges, which, thanks to Schulze's researches, 

 has been shown to be a connective tissue with distinct cellular elements, 

 fixed, wandering, contractile, glandular, &c. The presence of the 

 internal endothelium, absent in Protospongia ; and the fact that in 

 the latter the ciliated cells are almost wholly immersed in the con- 

 nective substance, are also noted. The impossibility of now denying 

 either the true sexual reproduction as evidenced by the repeated pre- 

 sence of spermatozoa, or the presence of two embryonic layers in the 

 larval forms is enforced. 



The metazoan nature of sponges is not however inconsistent with 

 their relation by direct descent from Choanflagellates, as Biitschli has 

 recently maintained, though such a history of the origin of sponges 

 would conflict with the other theory of their close relationship with 

 Cnidaria. Prof. Schulze reviews the opinions of Leuckart, Balfour, 

 Marshall and others in regard to the relation between sponges and 

 Coelenterates. Against Biitschli's hypothesis, he advances the apparent 

 absence of collared cells in the blastula stage, where they would, on 

 his supposition, be naturally looked for. He inclines towards the 

 supposition of an independent origin of the collars, hints of which 

 are found in some Protozoa, as in Placopus ruber. From the close 

 resemblance between sponge and Coelenterate larvae, Prof. Schulze 

 maintains that the divergence of the two lines did not begin before 

 that stage in the phylogenetic development, which corresponds to the 

 metamorphosis of the mature ciliated larvse. He thinks Marshall's 

 hypothesis that the common ancestors had radially disposed mesen- 

 terial pouches, tentacles with stinging capsules, and lateral pores, to 

 be without sufficient basis, and regards the most primitive type as 

 a simple sack-like form such as persists in Olynthus. 



Origin of new species owing to the loss of older characters.* 

 — The late Prof. O. Schmidt relates how in 1864 he was induced to 

 place Ancorina aaptos among the Tetractinellidai, owing to its 

 resemblance to the family of Corticate; this family he has since 

 had to regard as untenable, and with it A. aaptos had to be given 



* Zeitschr, f. Wiss. Zool., xlii. (1885) pp. 639-47 (1 pi.). 



