438 MM. D. C. Danielssen and J. Koren 
genus as a consequence of it. Our investigations lead us 
herein to agree with Dr. Viguier. It is true that in Solaster 
papposus the meshes of the calcareous reticulation are much 
larger than in S. endeca; but this can only come into consi- 
deration in specific determination. 
In Solaster affinis the meshes are smaller than in S. pap- 
posus; in S. furcifer they are still smaller; and in the new 
species, S. glacialis, established by us, the meshes in size ap- 
proach very closely to those of S. endeca. In all the reticulation 
is formed by larger or smaller, oblong or angular calcareous 
pieces, which are imbricated and form longer or shorter beams, 
by the union of which the reticulation is produced. 
The ventral interbrachial space in all the species named is 
occupied by calcareous plates, which are sometimes oval or 
nearly round, and contiguous without being truly imbricated, 
as, for example, in S. papposus and affinis; sometimes flatter, 
angular, and imbricated, as in S endeca, glacialis, and fur- 
cifer ; but these differences cannot serve as generic characters, 
nor has Viguier adopted them as such. 
The paxille, which are borne upon the calcareous reticula- 
tion, are placed closer together or further apart in the different 
species, generally according as the meshes are larger or smaller ; 
they are furthest apart in Solaster papposus, closest in S, en- 
deca. ‘Transitions occur distinctly in the three intermediate 
species, S. affinis, furcifer, and glacialis. As regards their 
form and development also we find transitions which may well 
assist in the discrimination of species, but certainly not of 
genera. ‘They are longest and have the longest pencil of 
calcareous needles in Solaster papposus; in S. affinis they are 
shorter, still shorter in S. furcifer and glacialis, and shortest 
of all in S. endeca. 
Prof. F. J. Bell, who has paid attention to the genus 
Solaster, is inclined to, support Dr. Viguier in his opinion 
against Agassiz’s division’; but he nevertheless expresses him- 
self with reserve, and says*, ‘‘ but that excellent investigator 
seems to me to have not fully weighed all the facts which can 
be made out with regard to these two forms, which he retains 
in one genus.’”’ He remarks further that the circumstance 
that Solaster endeca has only ventral and S. papposus only 
dorsal marginal plates may furnish very “ considerable sup- 
port” to Agassiz’s opinion. If this were really the case we 
should admit that there was some justification for making the 
division, although we cannot see that this peculiarity of the 
marginal plates alone was a sufficient ground for the division 
of the genus. Of the five species that were at our disposal 
* Ann. & Mag. Nat. Hist. ser. 5, vol. vin. p. 140 (1881). 
