764 ME. A. G. THACKER ON [June 16, 



Finally, a sponge described by Dendy [12] three years ago from 

 Ceylon under the name Leucosolenia tenuipilosa. is to be regarded 

 ■as a variety of this species and as standing in the same relationship 

 to typical specimens of L. canariensis as L. coriacea ceylonensis, 

 Dendy, does to the typical L. coriacea. This variety, L. canari- 

 ensis tentcijnlosa, has regular triradiates and quadriradiates with 

 rays 0*1 mm. long by 0*012 mm. thick, the apical x'ays of the 

 quadriradiates being very variable but usually longer than the 

 facial rays. The distinguishing feature of the variety is the 

 presence of hair-like oxea. Similar hair-like oxea occur in several 

 •of the Cape Verde Islands specimens under discussion. 



The collection contains a considei'able number of specimens 

 which I consider belong to this species. They were dredged by 

 Mr. Crossland fi'om various localities, Boa Vista Island, Porto 

 Praya, and elsewhere. The sponge forms rather massive colonies 

 of reticulating Ascon-tubes (Plate XL. fig. 3). The exhalent 

 openings are true oscula and there is no pseudoderm or endogastral 

 network, but in examples where the tubes anastomose very closely 

 more or less indefinite inter-canals are formed. Each colony 

 is attached by a somewhat constricted base, and in several cases 

 there is a short stalk. None of the colonies in my specimens is 

 very large ; the larger sponges have a diameter of from 1 cm. to 

 2 cm. The diameter of the Ascon-tubes varies considerably in 

 different specimens and in difierent pai-ts of the same specimen ; 

 this A^ariation ranges from 0"15 to 0'4 mm. The walls of the 

 tubes are from 0"02 to 0'035 mm. thick. The colour (in alcohol) 

 is brown. 



The Skeleton consists of regvilar triradiates, and of quadriradiates 

 which diflfer from the triradiates only in the presence of the 

 fourth ray ; some of the triradiates possess a knob evidently 

 representing an incipient fourth ray. The rays are usvially fairly 

 bluntly pointed (though there is some variation in this respect) 

 and are nearly cylindiical. The a,pical rays of the quadriradiates 

 project at right angles to the facial rays. There is not much 

 variation in the size of the spicules in the same individual, but 

 much difference between the spicules of different specimens. The 

 measurements of the spicules of the following series of five speci- 

 mens show how completely this series links up L. ca7iariensis 

 on the one hand and L. tenuipilosa and L. nanseni on tne otner 

 the sizes of the spicules of these sponges having beer: given 

 •above : — 



Speciinen R.N. 6. — Lengtn of rays, 0'06-0'07 mm. 



Thickness of rays, 0-006-0-007 mm. 

 Apical rays of quadriradiates straight and equal or nearly 

 equal in length to facial rays. 

 ^Specimen R.R. 8. — Length of rays, 0*08-0"09 mm. 



Thickness of rays, 0-007-0-008 mm. 

 Apical rays of quadriradiates straight and from one-half to 

 two-thirds the length of the facial rays. 



