244 Proceedings of the Royal Irish Academy. 



001 mm. Similar hard specimens were taken in the Caragh River, about 

 half a mile below Cai-agh Lake, Co. Kerry; but here again the spicules had 

 changed sliglitly, beiu;^ longer, more slender, and possessing longer points 

 than those of the lake specimens. In Connemaia, where llie lakes are often 

 united by channels, sometimes only a few yards in length, really typical 

 specimens of //. Kyderi were not found. In these short, though sometimes 

 rapid, streams, H. Eydcri was hard ; but, as in the foregoing cases, the spicules 

 had begun to change in shape. 



Dr. Annandale (3, p. 40 and ]>. I'lii) notes a similar change in the 

 external appearance of an 1 ndiau f ivsii-water sponge, CmTo^ongillu lapidosa, 

 according as it occurs in still or running water. In the former the sponge 

 grows on the under surface of stones, in small crusts, whicli liave a flat 

 surface, except where the oscola are raised on conical eminences. In running 

 water the sponge grows in broad sheets, which have a corrugated 

 surface. This resembles the chan;j;e in appcanmce of II. lii/dcri. On the 

 other hand, the Indian sponge is harder in tt-.xtuie in running water than it 

 is in the lake, the opposite being the case with H. liydcri. Apparently the 

 spicules do not differ in the lake and liver fonns. 



A form of II. liydcri intermediate l>etween the iiard lake form and the 

 typical river form is found in Irish lakes, where the conditions are apparently 

 unfavourable to robust growth, but yet where the sponge grows fairly abun- 

 dantly. Such lakes are Lough Ouler and Upper an<l Lower Lougli Lray, 

 io Co. Wicklow ; the C'oumgorra l^kes, Co. Waterford ; Lough E:.iglier, 

 Co. Kerrj- ; and I>ough Cunnel, Co. Mayo. These lakes, it may lie noted, 

 lie mostly at high altitudes for tliis country. With the exception of Lough 

 Cunnel, which is at 690 feel, they lie between 1,22-"* and 1,896 feet. 



E.xternally the 8ix>nge taken in the foregoing localities and in other 

 similar situations resembles the lake form, except lliat it spreiids over a 

 greater area, and its outline is not so circular. It is soft to the touch, but is 

 not so lobed as is the river form. Un the other hand, its spicules are similar 

 to those of the river form, from which they cannot be distinguislied 

 (I'l. XX VI 1 1, tig. 8). In some cases thei-e is, perliaps, a larger proportion 

 of straight spicules tiian in river specimens. Gemmules are, as a rule, 

 present in fair numbers, another j>oint of difference from tlie usual lake form. 

 In the report on the fresh-water sjwuges of the Clare Island Survey, the 

 Lough Cunnel sponge is referred to (,41 p. 14) as being of the typical form 

 oi U. Ryd/Tx. With material from otiier localities available for examination, 

 I do not now consider this sponge, and others similar to it, to be altogether 

 typical, and l>elieve that, strictly si>eaking, the typical H. /lyferi occurs solely 

 in runniug water, in streams and rivers. 



