from the North Sea. 83 



needle. Figs. 15 and 16 show beginnings of such branches 

 directed towards the middle of the needle; figs. 17 and 18 the 

 same directed towards the point. Sometimes the branching is 

 double crosswise, four branches with four canals (fig. 19), 

 sometimes regularly, sometimes rather irregularly, or in con- 

 nexion with bifurcation (fig. 20). I have, besides, several 

 times found an irregular heap of round, bladder-like tuberosi- 

 ties (figs. 21, 22, 23), to which the central canal gives no 

 branches. Often there are spicules with graduated points 

 (figs. 24, 25) ; very seldom their surface is studded with short, 

 pointed projections (fig. 26). 



When the spicule is perfectly entire and uninjured, the con- 

 tents of the central canal, even after boiling in nitric acid, re- 

 tain their transparency ; but if the spicule has been broken, 

 even scarcely perceptibly, at the outermost point, the canal is 

 partly filled with long, interrupted columns of gas, less trans- 

 parent than the lumen of the canal (figs. 28, 29, 30). 



Prof. Lieberktihn observed the first formation of siliceous 

 spicules in young individuals of Sjpongilla^. In a cell with 

 nucleus and nucleolus there appears among the granules a little 

 ball of silica, from which, in opposite directions, but not exactly 

 in the same straight line, shoot out two points, which are little 

 by little elongated, until they form spindle-shaped needles, the 

 ball remaining near the middle as the nodule. It is hardly to 

 be doubted that the inflation or nodule in the spindle-shaped 

 needles of our sponge, and which, as long as it is of small 

 size, receives no branches from the central canal, is the part 

 earliest formed — the siliceous ball. Of the growth of the 

 needle, free in the parenchyma, we know at present very little. 

 It increases by layers one over another. Prof. Kolliker, who 

 regards the canal as a solid fibre of soft organic matter, on 

 which, within the cell and from its contents, silica is deposited, 

 supposes that the spicule increases by secretion of silica from 

 the parenchyma in layers one above another f. In our sponge 

 these layers are scarcely discernible. But another siliceous 

 sponge from the Arctic Sea has offered some observations 

 which may deserve to be previously mentioned here. The 

 layers are very distinct, and seem to be alternately soft and 

 hard. A spicule has lost, near the end, its exterior layer, so 

 that the point projects beyond the remaining part of it, as out 

 of a sheath. Between the outermost broken lamella and the 

 exterior surface of the uninjured point there is a space, the 

 former contents of which, a soft substance, have disappeared, 

 the Canada balsam now occupying their place. If one of the 



* Muller's Archiv, 1856, p. 408, t. 15. f. 17-23. 

 t Icones histiologicae, i. p. 61. 



6* 



