360 ME. A. DENDT ON WEST-INDIAN CHALININE SPONGES. 



connecting-links between the variety in question and the spinose specimens, that it is 

 impossible to make a specific distinction between them. 



It is, however, very convenient to keep the two genera distinct, and I have there- 

 fore not united them 1 . 



The first species described by Duchassaing and Michelotti is " Tuba Sancta cruris ; " 

 but this is not even figured, and we have only a very meagre description of the external 

 form. The second, however, viz. Spinosella {Tuba) sororia, is recognizable with 

 tolerable certainty from the figure. 



Spinosella sororia, Duchassaing and Michelotti, sp. (Plate LVIII. fig. 7 ; 

 Plate LIX. fig. 1.) 



1864. Tuba sororia, Duchassaing and Michelotti, Spongiaires de ia Mer Carai'bs, p. 46, pi. viii. 



fig. 1. 

 1870. Siphonochalina papyracea, Schmidt, Spong. atlant. Gebiet. p. 33. 



This species is shown by the British Museum collection to be an extremely variable 

 one, and it is very probable that several of the forms described by Duchassaing 

 and Michelotti as distinct should be united with it, as already pointed out by Schmidt 

 (loc. cit.) in his valuable remarks on the subject. 



I cannot, however, agree with Schmidt in suppressing Duchassaing and Michelotti's 

 name and giving a new one, viz. Sijphonochalina papyracea, to a species which has 

 already received far too many. I have therefore retained the specific name sororia, 

 which represents the first recognizable species of Duchassaing and Michelotti's genus 

 Tuba, and propose to describe in this place four well-marked varieties of the species, 

 all of which agree so closely with one another in microscopical structure and are con- 

 nected by so many links that it is impossible to separate them specifically. 



The typical form (PI. LIX. fig. 1), agreeing closely with Duchassaing and Michelotti's 

 original figure, is represented by several specimens. The one which I have selected for 

 description consists of a number of long tubes united together in an irregular basal 

 mass. The whole sponge is compressed and fan-like, but this condition may be 

 partly due to artificial pressure. The tubes vary in length from 8 to 16 centini., and 

 in greater diameter from 2 to 3-5 centim. ; they may be entirely free from one another 

 except at the base, or more or less united laterally. The inner surface of the tubes is 

 smooth and appears strongly veined, the venation being due to an unusually strong 

 development of the main skeleton reticulation in the places where it occurs. The outer 

 surface is strongly spined. The spines average about 7 millim. in length, and are caused 

 by projections given off obliquely outwards and upwards from the longitudinal veins just 

 mentioned. Around the margin of the tubes the ends of the veins project freely, and 

 cause the orifice to appear " ciliated." The true oscula are small and circular, and 

 abundantly scattered on the inner surface of the tubes. 



1 Cf. Ridley and Dendy, Report on the Monaxonida collected by H.H.S. ' Challenger,' p. 29. 



