1904.] SPONGES OF THE GENUS LEUCOSOtENIA. 375 



across under the middle of the sponge-colony, which was thus 

 detached in two halves and somewhat daiiiaged, but still showing 

 well the peculiarities of this mode of growth. The spiculation of 

 this specimen was foiuad to be normal in all respects. My second 

 specimen of the spreading form was removed entire from the rock 

 on which it grew ; it is a small colony i-emarkable for the close 

 network of basal tubes, giving the sponge almost the appearance 

 of a Glathrina, wei-e it not for the characteristically lai'ge oscuLu- 

 tubes. The spiculation of this specimen shows a comparative 

 scarcity of monaxons and unusually small triradiate systems ; in 

 other respects, however, the charactei-s are typical (text-fig. 95, 

 figg. 14 «-14/, p. 379). Johnston [14, pi. xxi. fig, 3] has figiired 

 a typical spreading specimen of variabilis under the name of 

 Grantia hotryoides var. hmiantia, which Haeckel has wrongly 

 placed as a synonym of Clathrina coriacea. 



L. variabilis occurs most commonly in the bushy form as a 

 compact i-eticulum of fine anastomosing basal tubes from whicli 

 arise the stouter oscular tubes, often closely packed and of con^ 

 siderable length. In this form it may be found in rock-pools 

 attached between the stems of algse, or ci-eeping over the algfe 

 themselves. In the former situation the basal portion of the 

 sponge is often half buried in mud and sediment, and, doubtless 

 in consequence of this, the oscular tubes grow to a great size and 

 length. I have such a specimen from Plymouth, in which the 

 oscular tubes average 1'5 cm. in length, reaching in some cases 

 2 cm. ; the spiculation of this specimen (text-fig. 95, figg. 15 a-l5fi 

 p. 379) is remai"kable for the large size of the triradiate systems. 

 It is more usual, however, for the sponge to form a creeping 

 growth over the algee themselves — either twining amongst the 

 filaments of confervee, as Haeckel's fig. 6 on pi. 18 of the ' Mono- 

 graphie ' [13] shows fairly correctly, or spreading over the stems 

 of stouter seaweeds. These creeping forms are commonly found 

 at Roscoff in situations where they are left dry at all tides, growing 

 amongst the stems of the dense growth of algse covering isolated 

 boulders on the seashore. The youngest colonies form a delicate 

 network spreading over the seaweed stems and sending up oscular 

 tubes at intervals, thus having a form similar to the spreading 

 colonies already described ; but with further growth the basal 

 tubes form a compact tangled mass from which long osculai' tubes 

 arise or hang down as the case may be. 



In all the modifications of form exhibited by this sponge the 

 m.ost constant feature is to be found in the oscvilar tubes, which 

 show little or no tendency to throw out diverticula except near 

 their base. Hence the sponge shows numerous oscular tubes 

 arising from the basal reticulum, each long, slender, geneiully 

 slightly curved and smooth, i. e. fi'ee from diverticula from near 

 their base up to the oscular opening. We find the sharpest con- 

 trast, as has already been stated above, with L. complicata in this 

 respect, and the typical arborescent form of the latter is never 

 found in L. variabilis. 



