sind als Spiraster vorhanden, welche eine besondere Rindenschicht erzeugen können. Übergange 

 von Spiraster zu Aster bisweilen vorhanden. Kanalsystem nach dem vierten Typus". In 1897 

 (p. 217) Lendenfeld again modified the diagnosis: "Massig"e oder Krusten-bildende, freie, nicht 

 bohrende Spirastrellidae mit einfacher, nicht wabiger oder papillöser Oberflache und glattrandigen 

 Osculis. Mit tylostylen Megascleren und Spirastern, dornigen Microrhabden, oder Amphiastern". 

 Topsent (1898 |3 p. 100 and 1900 p. 107) adopts on the whole the definition of Ridley & 

 Dendy adding, however, that the megascleres sometimes are "tornotes". With exception of the 

 latter clausule most authors agree with him. No one, however, till now has unclertaken to 

 enumerate which sponges described by previous authors should be received by the genus 

 Spirastrella. Nor do we find much about the anatomy of single species, still less about com- 

 parative anatomy. Apart from good (or bad) descriptions of the skeleton and its spicules, we 

 hardly find anything about the soft parts and the canalsystem. Whereas Keller says that (in 

 S. deaimbens) it is built "nach dem vierten Typus", Lendenfeld states that (in 6". bistcllatd) 

 the mastichorions are "kuofligr oder oval" and that "sie münden direkt mit weiter Offnunsf in 

 die ausführenden Kanale, welche ungefahr ebensoweit als die einführenden sind". Any distinct 

 illustration, demonstrating how the author supposes that things are, is absolutely wanting. I 

 thought it, therefore, useful to try to fill the gap since I had such a rich material at my 

 disposal as the Siboga brought to light. It has shown me once more how valuable it is to 

 collect as many specimens as possible — as will be shown hereafter. But I wish to draw 

 attention also to another fact, which I strongly recommend future collectors to take to heart: 

 I mean the necessity of not only preserving many specimens in the strongest alcohol, but also 

 of rapidly drying specimens. The majority of sponges we find in our musea are in the dry 

 state. For identification it is often necessary to be able to compare them with likewise dried 

 specimens. The external aspect may undergo such changes, that a prima vista a dried and an 

 alcohol-specimen sometimes hardly resemble each other. 



As far as I can judge I found in the Siboga collection 40 or 50 bottles, containing 

 together about 80 specimens, which I believe to be Spirastrellae ; in addition there are about 

 ten dry specimens. Some of these I thought I could identify with sponges described by former 

 authors: Bowerbank, Carter, Dendy a.o. At a closer examination I found, however, considerable 

 differences in spiculation, where the external appearance showed the most striking- resemblance 

 and vice versa. I thought it advisable in order to settle the question, to. pay a visit to the 

 British Museum. I feel greatly indebted to my friends S. F. Harmer and R. Kirkpatrick for the 

 kind way they received me and the great help they gave me in putting the wanted specimens 

 at my disposal. Nor shall I forget my visit to the nestor of spongiologists, my old friend 

 Canon A. Merle Norman in Birkhamstead, to whose liberality I owe several valuable prepa- 

 rations. The result has been, that in most cases I could consult the type specimens and compare 

 them with the Siboga Collection. All in all I carefully examined more than 125 specimens; 

 I arrived at the conclusion that, with a few exceptions, they represent but one single 

 species. Indeed an exceedingly variable species. A priori nobody would have said so who 

 saw the various specimens ; not even I who again and again arrived at the conclusion that 

 sponges vary ad infmitum. As is shown in the lists, given hereafter, there are, described, 



