55 
Sammenvoxningen er saa komplet, 
den. 
Randplader fandtes. 
at vi ikke med kaustisk Kalilud have kunnet hve 
De ventrale Randplader, Tab. IX, Fig. 12, b, ere 
næsten lancetformige, deres bredere Del vender udad, og 
dennes undre Flade afgiver et stærkt Støttepunkt for de 
dorsale Randplader. Den indre, noget smalere Del, har 
en konkav Overflade, der hviler paa den undre, afrundede 
Rand af Adambulacralpladerne, Fig. 12, a; den undre 
Flade er ujævn konvex. De ventrale Randplader bære 
ingen Paxiller, men tjene udelukkende som Støtte for de 
dorsale Randplader. Det falder af sig selv, at i Inter- 
braehialrummene støde de ventrale Randplader umiddelbart 
til Bugpladerne. 
Odontophoren (Interbrachialpladen, Agassiz) frembyder 
ingen nævneværdig Forskjellighed, men har saavel hos Sol. 
papposus som endeca den for Slægten Solaster angivne Type. 
Det egentlige Skelet er saagodtsom ens for begge 
Former. Saavel Ambulacral- som Adambuleralpladerne staa 
kanske lidt længere fra hverandre hos Sol. papposus, end 
hos endeca; men ikke engang denne lille Forskjellighed er 
konstant. Tænderne ere noget stærkere hos Sol. endeca, 
end hos papposus; men forresten ere de ens. 
Efter hvad vi have fremhævet af disse to Arters Ana- 
tomi, fremgaar det formentlig med Bestemthed, at Forskjel- 
lighederne ikke ere større, end de nødvendigvis maa være, for 
at Arterne skulle kunne opretholdes som saadanne, og at 
der, ifølge vor Opfatning, ikke bør være Tale om at dele 
Slægten for de omhandlede to Arters Vedkommende. 
Skulde en slig Deling finde Sted, blev det nødvendigt at 
danne en egen Nlægt ogsaa for vor nye Art, Sol. glacialis; 
thi den kan efter den Maalestok hverken henføres til Sol- 
aster eller Crossaster, saasom den staar omtrent midt imel- 
lem. Den danner en Overgang fra Sol. endeca til papposus. 
Man vil se, at Paxillerne hos Sol. glacialis staa læn- 
gere fra hverandre og ere lidt længere end hos endeca, 
men ere hverken saa lange, eller saa spredte som hos pap- 
posus; at der langs Armenes Rand er en Række temmelig 
store, penselformige Paxiller, der sees tydeligst fra Rygsi- 
den ligesom hos papposus, uden at fremtræde saa stærkt 
som hos denne; imedens de paatversgaaende Paxillerækker, 
der ere saa egne for Bugfladen hos Sol. endeca, langtfra 
ere saa fremtrædende hos Sol. glacialis. Og ser man nu 
hen til Hudskelettet, saa finder man, at de smaa Kalk- 
stykker, der danne Maskenettet, have i Form og Udbred- 
ning meget tilfælles med det hos endeca; men mest nær- 
mer det sig dog Sol. furcifer, som det med Hensyn til 
Randpladernes Anordning og Form ligner ganske. Vi for- 
mene saaledes, at Sol. glacialis viser end yderligere Uhold- 
barheden af den omtalte Deling af Slægten Solaster. 
found. The growing together is so complete that a solu- 
tion of caustic potash, has been unable to effect a sep- 
aration. 
The ventro-marginal plates (Pl. IX, fig. 12, b) are 
almost lancet-formed, with the broad part turned outwards, 
and the inferior surface affords a strong supporting point 
for the dorso-marginal plates. The and somewhat 
narrower part, has a concave surface reposing on the inf- 
erior rounded margin of the adambulacral plates (Pl. IX, 
fig. 12, a). The 
ventro-marginal plates carry no paxille, but serve exclus- 
ively, as supports for the dorso-marginal plates. It follows, 
naturally, that, in the interbrachial spaces, the ventro- 
marginal plates abut close upon the ventral plates. | 
The odontophore (the interbrachial plate, — Agassiz) 
presents no noteworthy divergence, but has, in Solaster 
papposus, as well as, also, in Solaster endeca the normal 
type pertaining to the genus Solaster. 
The skeleton proper is almost exactly alike in both 
forms. The ambulacral plates, and also the adambulaeral 
plates, are, perhaps, placed a little further apart from each 
other in Solaster papposus, than in endeca, but even this 
small divergence is not constant. The teeth are somewhat 
stronger in Solaster endeca, than in papposus, but otherwise 
they are alike. 
From what we have now particularised in reference 
to the anatomy of these two species, it may, we think, be 
accepted as certain, that the divergencies are not greater 
than they of necessity must be, in order to permit the 
inner 
The inferior surface is roughly convex. 
species to be maintained; and according to our view, there 
ought to be no question of subdividing the genus, as far 
Should such a 
subdivision be made, it would also become necessary, to 
form a special genus for our new species Solaster glacialis, 
because, it can, according to this same scale, be assigned 
neither to Solaster nor Crossaster, inasmuch, as it stands 
about intermediately between them, and forms a transition 
from Solaster endeca to Solaster papposus. 
It will be seen; that the paxille in Solaster glacialis 
are situated more apart from each other, besides being a 
little longer than in endeca; but are neither so long, nor 
so dispersed as in papposus; further, that along the margin 
as regards the two species referred to. 
of the rays, there occurs a series of rather large penicilli- 
form paxillæ, which are seen distinctly from the abactinal 
side; and this, is the case, also, in Solaster papposus, but they 
are, Inendeca, not so prominent as in papposus, whilst, the 
transverse paxillar series which are so peculiar to the 
actinal surface of Solaster endeca, are not nearly so prom- 
inent in Solaster glacialis. If we proceed to look at the 
dermal skeleton, we find that the small calcareous ossicles 
which form the reticulation, are, in distribution and form, 
very similar to those of endeca, but yet, approach still more 
closely, in similarity, to Solaster furcifer, and indeed, so much 
so, in regard to the form and arrangement of the marginal 
plates, that these are quite alike. We think therefore that Sol- 
aster glacialis still further shows the untenability of the sub- 
division of the genus Solaster, which has been referred to. 
