﻿135 "ENDEAVOUR" SCIENTIFIC EESULTS. 



(ii.) The amphiastersi bear at either end a terminal whorl of 

 about six to eight rays which stand almost perpen- 

 dicularly to the shaft or are directed forwards at a 

 small angle (up to 30°, say). The width of the spicule, 

 between the tips of the rays, measures from about 

 three-fourths of, to slightly less than the length of 

 the spicule, which is rarely as much as 5 p. 



Loc. — Coast of New South W'ales, eight miles east of San- 

 don Bluffs, 35-40 fms. ("Endea\our. ") 



Family DESMACII^ONID.'E. 

 Whether the Desmacidonida^ be divided into two sub- 

 families, Mycalinae and Ectyoninae, in the usual way ; or 

 whether, in accordance with Topsent's proposal, the Myca- 

 lina? be sub-divided so as to yield an additional sub-family, the 

 Dendoricinae — the result, if a natural grouping of the genera 

 be our aim, is not altogether satisfactory. Of the two, Top- 

 sent's classification is the better since, with very few excep- 

 tions as their microscleres show, the genera which remain in 

 the Mycalinae after the removal of the Dendoricinae stand well 

 apart from the rest of the family. The main fault lies in the 

 artificiality of the distinction upon which the separation of the 

 Ectyonina? and Dendoricina is based. Species are placed in 

 one or the other of these sub-families according as spicules of 

 one particular kind — the accessory spicules — are present or 

 absent, whilst no value at all is placed upon the presence or 

 absence of the equally important "skeletal" and "dermal" 

 spicules, nor any notice taken of the plain indications afforded 

 by the microscleres. As a consequence, we have closely allied 

 if not almost identical genera like Ectyodoryx and Lissoden- 

 doryx, Eciyoniyxilla and Myxilla, Pocillon and lophou, Hyme- 

 trochota and lotrochota placed, the one in the Ectyoninte, the 

 other in the DendoriciucC, whilst on the other hand wholly un- 

 related species like those of JViJsoneUa and Claihria are put in 

 a single genus without question or comment. Under these 

 circumstances there is no advantage in maintaining these sub- 

 divisions, and I propose therefore to merge the Dendoricin.-e 

 and Ectyoniuce in a single sub-family, to be called the Myxil- 

 lina?. It is probable that the Myxillinae are capable of sub- 

 di\'ision into two fairly natural groups in the wav suggested 

 below, but on this question I prefer to reserve judgment. I 

 would, however, \'enture the opinion that the Myxillinae are 

 derived from a single stem distinct from that from which the 

 great majority of the Mycalinae have sprung, and from this 

 point of view will enter upon a brief discussion of the sub- 

 family. 



1 Not having succeeded in making out the details of the structure of 

 this spicule, I have not attempted to flg'ure it. The rays appear to be 

 often bifurcate and to have at times a slightly tuberculated surface. 



