3. Calcarea. 13 



believes the cells of the nascent gemmule to be identical with those cells of the 

 parenchym that similarly contain Zoochlorellae and unequal granules. The green 

 colour is masked as the yellow-brown cuticle develOpes. The green gemmules 

 form green sponges, the white gemmules colourless, grey, or brown sponges. In 

 Eu. lacustris the old gemmules after December show yellowish-green Zoochlorellae, 

 the sponges growing from them have the same colour , and only become dark- 

 green later. Starch was not found in the chloroblasts in any Spongillid, certain 

 bodies of unknown nature occur. Each Zoochlorella possesses 1 to 3 chlorophyll- 

 bodies and clear protoplasm with bright granules. A small sponge growing in the 

 aquarium will gradually diminish in size, leaving a green trace on the glass, which 

 is found to consist of Protozoa, spicules, and free Zoochlorellae which have escaped 

 from the dying cells. Jackson says rightly that the inner part of green sponges 

 can be white, this agrees with what is known of algae in other animals. The 

 brown pigment of Sp. fragilis is due to angular yellow bodies 0,0008 mm to 

 0,0043 mm in diameter, retaining their colour in alcohol ; they are in the imequally- 

 granuled amoeboid cells and may quite fill them , again Zoochlorellae may occur 

 in the same cells. They are not an antecedent stage of the green granules; they 

 may be excretions of the cells ; they resemble some of the pigment granules in 

 Hydra fusca. They do not appear essential, not being present in sponges living 

 under stones etc. Influence of temperature on gemmules. The same 

 species produces gemmules at different times in different localities. W. has not 

 found the flagellate chambers and canals disappear in winter ; they diminish in 

 number. Gemmules developed into sponges after being respectively: - - (1) 17 days 

 frozen; (2) 59 days frozen with intervening thaws; (3) several days frozen, one 

 month dry, 18 days under snow; (4) 3 months in a hot room; (5) frozen, dried, 

 cut in two halves ; each half forming a young sponge. Gemmules which had been 

 dry 30 years failed to develope. Deformed larvae in which by violent treat- 

 ment the ectoderm had been injured or partly torn off, developed into young 

 sponges. 



3. Calcarea. 



See also Dendy ( 2 , 3 ) , Hartog, Lendenfeld, and supra p 8 Levinsen, p 9 Mar- 

 shall, p 2 Dendy (') 



Dendy (*) repeats his classification of the Heterocoela [cf. Bericht f. 1892 

 Porif. p 8], dropping the genus Paraleucilla. A detailed analysis is added of the 

 actual relations and hypothetical evolution of the different families and genera. 

 Leucascus stands apart, its skeleton has not yet attained to the typical radiate 

 condition of the Heterocoela , but retains the irregular scattered character of the 

 reticulate Homocoela. From these it appears to have been independently deri- 

 ved ; its flagellate chambers are long, irregular, and branching, there is no cloacal 

 cavity. It is separated from the Homocoela only by its distinct, pore-bearing, 

 dermal membrane, by the absence of collared cells from the exhalant canals, and 

 by the marked tendency of its flagellated chambers to become radial [but with 

 this compare Homocoela Radiata where from a single wide central tube . . . 

 numerous radially arranged branches are given off'f (p 247)]. Excluding L., all 

 Heterocoela are derived from the Sycetta form. Supposing the radial tubes of S. 

 more or less to adhere, and crowning each with a distal tuft of spicules, we obtain 

 a Sycon. In this genus is occasionally present the rudiment of a thin dermal mem- 

 brane. Grantia differs in possessing such a membrane strongly provided with 

 spicules ; and with all or most species in the presence of unordered radial oxeotes 

 (aeerates). An enormous development of cortex (5 mm. thick) distinguishes the 

 subgenus Gmntiopsis, while a cortex, supported by huge longitudinal oxeotes, 



