382 PROF. E. A. MIXCHIN ON THE Bpaxisii [Dec. 13, 



(2) .l.s reyards the form and size of the Spicules. — The ordinai'-y 

 ti'iradiates and qxiadi-iradiates vary considerably in size both in 

 diiFerent and in the saine specimens, though here, as in other 

 species, one specimen may be found characterised by spicules 

 generally above the average in size, another by spicules below 

 the avei-age. The smallest triradiates I have observed, occurring 

 in a specimen from Roscoff, have the unpaired rays abotit 60 /u, 

 the paired rays about 70 fx in length (text-fig. 95, fig. 14, p. 379). 

 The largest I have seen have unpaired rays 95-100^, paired rays 

 135-140 fji, in length (text-fig. 95, fig. 15). In thickness also they 

 vary considerably, ranging from 4-10 /x (text-fig. 96, fig. 21a, 

 p. 381, text-fig. 95, figg. 15 &, 15 c, p. 379), and, in the latter case, 

 only differing from the T-shaped forms by the paired angles and 

 the even curvature of the paired rays. On the other hand, as 

 regards foi'm and proportions, the triradiates are remarkably 

 uniform, the unpaired ray being almost invariably shorter than 

 the paired rays, usually very obviousl}^ so, and only very excep- 

 tionally eqvialling them in length. 



The bi-ackets are very variable both as regards the form and 

 degi-ee of development of the unpaired ray, and as regards the 

 length of the gastral ray when developed ; the latter always 

 greatly exceeds in length, howevei-, the gasti-al rays of the ordinary 

 quadriradiates . 



The cvirved monaxons {b, 1) are extremely variable in size. 

 The normal condition is for every intermediate grade to occur 

 between spicules of about 80 fx in length and others three times as 

 long. In some specimens, however, only small monaxons occur, 

 in others the majority of them are large, reaching nearly 400 /.t in 

 length and correspondingly thickened ; but in the latter case 

 small forms are always present in addition to the large, though 

 the}^ may be scarce by comparison. 



The aberrant specimens of L. variabilis from Brighton Aqua- 

 rium, which were described by Bowerbank under the name 

 Leuconia somesii (text-fig. 95, figg, 18, 19, p. 379), are characterised 

 by an extraordinary development of the monaxons, which, in 

 additioia to ordinary small and large forms, reach in other cases a 

 great length (as much as lOOO/x) combined with unusvial slender- 

 ness foi- such a length (2 5 /x). The triradiates of these specimens 

 are also characterised by great range of variation in size, and 

 by the large number of abnormalities in this class of spicules. 



In certain specimens of this sponge which I have been able to 

 examine — specimens important for the i-eason that they had all 

 been through Haeckel's hands — I was puzzled by finding certain 

 types of spicules in my preparations in addition to those already 

 described (text-fig. 96, figg. 21 ^--21^^). The specimens in question 

 were, fii-st, four in Norman's collection (Nos. 1-4 of my list 

 below) ; secondly, one in the Berlin Museum (No. 1 3 of my 

 list) (text- fig. 96, fig. 21, p. 381), All these five specimens 

 show the typical spiculation of Leucosolenia variabilis, the 

 specimen from Polperro (No. 5) alone showing an admixture of 



