410 
Magellanic Echinoderm faunas. To these must also be added some 
few cases of nearly related species occurring in the two regions, 
viz. Notechinus novæ-zealandiæ—magellanicus, Amphioplus basilicus 
—texiilis, Allostichaster insignis — æqualis, Amphiura eugeniæ — var. 
latisquama. 
If we regard more closely into the matter, we find that, excepting 
Gorgonocephalus chilensis, all the identical or nearly related species 
occurring in the two regions live in the littoral region and are 
thus very liable to be carried along in the roots of large floating 
algæ. Gorgonocephalus chilensis is not known to occur in the littoral 
region and therefore a transport on floating algæ cannot explain 
its occurrence both in the Magellanic region and at New Zealand ; 
but it may well be assumed to have free swimming larvæ, which 
would easily account for its occurrence in both regions, the more 
so at it is known also from Kerguelen. 
Thus, we need not h a v e r e co u r s e t o a n y fo r m e r 1 a n d 
or shallow water connection between New Zealand 
and S. America (Patagonia) in order to explain the simi- 
larity in their Echinoderm faunas. It may rather be main- 
tained, on the contrary, that the Echinoderm faunas of the two 
regions bear witness against such former connection in post-mesozoic 
times. If such had existed, it should be expected that the Ophio- 
notus- and the Abatus-gvou^i, so highly characteristic of the Magellanic 
region, would be represented also in New Zealand, especially at 
the Auckland-Campbell Islands. But this they are, evidently, not; 
otherwise I could hardly have failed to discover them there in the 
several dredgings which I undertook, especially with the object of 
ascertaining their existence or non-existence there. Both these forms 
belong to the deeper littoral, and a transport over the open sea 
on floating algæ evidently is out of question; moreover they are 
viviparous (excepting Ophionotus victoriæ). They are, therefore, of 
the highest zoogeographical importance, their occurrence in distant 
localities (e. g. S. America and Kerguelen) giving strong evidence 
of former direct connection between such localities. As both types 
are widely distributed in the Antarctic and Subantarctic regions, the 
conclusion seems justified that, if there had been a former land- 
connection between the Auckland-Campbell Islands and the Magel¬ 
lanic region, or with Antarctica, one should expect to find both 
