The only case besides Spongilla in which the details of development from gemmules have been traced is that of Tethya.[[259]] Mere microscopic examination of a Tethya in active reproduction would suggest that the process was simple budding, but Maas has shown that the offspring arise from groups of archaeocytes in the cortex, that is to say, they are typical gemmules. As they develop they migrate outwards along the radial spicule-bundles and are finally freed, like the buds of the Hexactinellid Lophocalyx (Fig. 113).
The comparison of the process of development on the one hand by gemmules, and on the other by larval development, is of some interest.[[260]] In both cases two cell layers—a dermal and a gastral—are established before the young sponge has reached a functional state. Differences of detail in the formation of the chambers occur in the gemmule; these find parallels in the differences in the same process exhibited by the larvae of various groups of sponges. On the other hand, the order of tissue differentiation is not the same in the gemmule as in the larva.
Fig. 116.—Development of the triradiate and quadriradiate spicules of Clathrina. (1) Three scleroblasts; (2) each has divided: the spicule is seen in their midst; (3) addition of the fourth ray by a porocyte. p, Dermal aperture of pore; r, fourth ray. (After Minchin.)
Of the reproduction of Tetractinellida extremely little is known. Spermatozoa occur in the tissues in profusion and are doubtless functional, but larvae have been seldom observed.
Fig. 117.—Three stages in the development of the triradiate spicules of Sycon setosum. × 1200. (After Maas.)
In Hexactinellida the place of sexually produced larvae is taken by bodies of similar origin to gemmules but with the appearance of parenchymulae. Ijima has indeed seen a few egg-cells in Hexactinellids.[[261]] He finds, however, that archaeocyte congeries occur in abundance, and there is good reason to believe with him that these are responsible for the numerous parenchymula-like asexually produced larvae he has observed. The discovery of "asexual larvae" was first made by Wilson in the Monaxonid Esperella; in this case the asexual larva is, as far as can be detected, identical with that developed from the fertilised egg. A similar phenomenon, the production of apparently identical larvae by both sexual and asexual methods, has been observed in the Coelenterate Gonionema murbachii.[[262]]
Artificially, sponges may be reproduced with great advantage to commerce by means of cuttings. Cuttings of the bath sponge are fit to gather after a seven years' growth.
The development of the various forms of spicules is a subject about which little is yet known. Most spicules of which the development has been traced originate in a single dermal cell. The triradiate and quadriradiate spicules of Homocoela (Clathrinidae), as Minchin[[263]] has most beautifully shown, form an exception. Three cells co-operate to form the triradiate; these three divide to give six before the growth of the spicule is complete. A quadriradiate is formed from a triradiate spicule by addition of the fourth ray, which, again, has a separate origin in an independent cell, in fact a porocyte. The triradiate spicules of the Sycettidae, on the other hand, originate in a single cell,[[264]] but the quadriradiate spicules are formed from these by the addition of a fourth ray in a manner similar to that which has just been described for Clathrinidae.