Potash on the Siliceous Skeletons of Sponges. 293 



globo-stellate *, leaving only a thin outer film, perforated in 

 places and much resembling a broken egg-shell. If Bowerbank 

 had under his observation globo-stellates which had suffered this 

 change under the influence of atmospheric solvents, he might 

 well be forgiven for falling into the error he did respecting 

 them. 



By careful observation, It will in most cases be found that the 

 potash obtains entrance to the interior of these spicules by way of 

 the"hilum" which characterizesallof them. One can sometimes 

 look down and through the hilum into the cavity excavated 

 within; and frequently one may observe by a lateral view a canal 

 from the hilum in direct continuity with the inner cavity (PI. 

 IX. figs. 8, 9, 11). It is very noteworthy that while solution 

 has proceeded to such an extreme extent in the interior of the 

 globo-stellates, the exterior is, on the other hand, often scarcely 

 affected. Originally the outer surface is ornamented by spines, 

 which divide at their summits into four or five smaller spines, 

 and which are united together by lateral ridges into a compact 

 network ; the chief result of the action of the potash on the 

 exterior is apparently to dissolve away these connecting ridges, 

 and so to leave the spines in isolation and much better exposed 

 for an examination into their minute characters (PI. IX. figs. 

 12, 13). The reason for the difference in the rates of solu- 

 tion within and without these structures appears to lie in the 

 fact that when once the potash has reached the centre of a 

 globo-stellate, it is able to attack the ends of the rays of which 

 it is composed and to penetrate between their sides, while it 

 finds it more difficult, apparently impossible, to insinuate itself 

 between them when acting on the outside ; thus the surface 

 over which the potash in the interior acts is many times larger 

 tlian that exposed to the influence of that without, and in 

 consequence the internal solution is necessarily much more 

 rapid. In some few cases I have observed globo-stellates with 

 all their spines removed from the exterior, so as to present a 

 smooth surface : in these the way to the interior appears to 



* It often happens that the central cavity of such globo-stellates con- 

 tains a white opaque kernel-like spherule (PI. IX. fig. 7, s), the nature 

 of which was at first very puzzling to me ; for though I suspected it to 

 be an air-bubble, I could get no metallic lustre from its surface with 

 reflected light. By pressing on the covering-glass immediately over one 

 of these spherules, however, with a pointed piece of wood, it was found 

 possible to crush the globo-stellate containing it ; for the globo-stellates 

 when partly dissolved are very brittle and break easily. There then 

 remained only transparent fragments of the broken shell and an indubita- 

 ble glistening air-bubble lying in their midst. By carefully performing 

 the operation one could, indeed, see the air-bubble issue froin the globo- 

 stellate at the moment of fracture. 



Ann. & Mag. N. Hist. Ser 4. Vol. xx. 21 



