19 



fairly closely with one another in external form", we may take it for granted that this N" 60 

 is the co-type, named by Dendy. 



2. .V. inconstans var. maeandrina. Dendy's type-specimen, figured by him (1. c. PI. X, hg. 1) 

 is in the British Museum, registered '87.8.4. i; it bears my number B.M. 9 Dendy speaks 

 quite definitely of "the single specimen". Consequently is the specimen in the Madras Museum 

 to which I have given the number B. M. 59 and which seems to be named "Suberites inconstans 

 Dendy. Var. maeandrina. Pamban", not determined by Dendy. We were unable to make out 

 who did it. 



3. S. inconstans var. digitata. The type-specimen, figured by Dendy on PI. IX, fig. 2, 

 is registered in the British Museum 'S~. 8.4. 3, corresponding to my number B.M. 23- Dendy 

 says (1. c. p. 155): "There are three specimens which I refer to this variety (viz. digitata). 

 They difier considerably from one another in external appearance, but all of them show a more 

 or less strongly marked tendency to form digitate processes". Now there is another specimen 

 in the British Museum, registered '87. 8. 4. 4, which I will designate with B.M. 11 (Pk II, fig. 2). 

 Kirkpatrick writes to me that this is probably the co-type of which Dendy writes that "the 

 digitate processes are almost obsolete". I am willing to accept Kirkpatrick's suggestion. 



The second co-type is probably the one I mark M.M. 62 (Pk IV, tig. 1). About this 

 specimen I learn by letter from Kirkpatrick: "The specimen of digitata returned to Madras 



is almost certainly N° 62 the one of the Museum three of which Dendy says: 'one has 



the digitate processes very broad and irregular and with a very uneven corrugateel surface'. 

 Now the other two Madras digitata specimens (according to the photo's) have a smooth surface". 

 Again in this respect I agree with Kirkpatrick. 



Thiele described in 1899 some specimens of a very variable sponge, in which he found, 

 in addition to tylostyles, small spinispirae. He is of opinion, that his sponges are identical with 

 Dendy's Suberites inconstans and proposes, consequently, the name Spirastrella inconstans (Dy.). 

 Finally, in 1900, Thiele suggests the identity of S. inconstans and S. vagabnnda. In fact there is 

 much to say in favour of this suggestion ; many specimens of the Siboga-material form transitions 

 between vagabnnda and inconstans. Dendy's Suberites inconstans var. digitata likewise forms a 

 transition between inconstans and vagabnnda. Dendy says of it (1. c. p. 155): "They differ 

 considerably from one another in external appearance, but all of them show a more or less 

 strongly marked tendency to form digitate processes". Whereas the type (B.M. 23) possesses 

 long and clistinctly digitate processes, one of the co-types (B.M. 62) "has the digitate processes 

 very broad and irregular", while in the other (B.M. 11) "the digitate processes are almost 

 obsolete". The difterence between Dendy's fig. 2 on PI. IX representing B.M. 23 and my 

 illustrations PI. II, fig. 2 (B.M. 11) and PI. IV, fig. 1 (B.M. 62) are very considerable indeed. 

 Now Dendy writes (1905 p. 123) about Spirastrella vagabnnda var. tnbnlo-digitata: "In this 

 variety the sponge consists of hollow, finger-shaped processes or 'fistulae' rising from a sandy 



base " About S. vagabnnda var. trinconialioisis Dendy says: "It consists of a massive 



base rising up into a few short, stout, finger-shaped processes". To some of the Siboga 

 specimens the description of .S'. vagabnnda is just as well applicable as that of S. inconstans ; 

 e. g. : 638, 1103, 1211 (PI- II, fig- 4)1 1260, 1610 etc. In 1341a (PI. II, fig. 3) one or two processes 



