36 THE ECHINODERMS OF TORRES STRAIT. 
OPHIDIASTERID£E.: 
As already mentioned, the Ophidiasteridz are by far the most numerous and conspicu- 
ous family of sea-stars on the reefs at Mer. Since 10 of the 20 genera which comprise the 
family are found in Torres Strait and 7 others are represented in the Museum of Compara- 
tive Zodlogy collection, it seems appropriate to give here a complete revision of the family. 
A very useful key to the genera was published by Sladen (1889) and this has been revised 
and improved by Fisher (1911), but the key here offered (p. 37) includes 9 genera not 
recognized by those writers. 
The Ophidiasteride are essentially a tropical, shallow-water family and very few 
species are known from outside the tropics or from water over 20 fathoms deep. Most of 
them occur among rock fragments and covals on reef flats or along shore, but large individ- 
uals often lie exposed on sandy bottoms or on the surface of the reef. Most species have 
remarkably small disks and long cylindrical arms, but trigonal arms are fairly frequent 
and in Fromia and Ferdina the disk is often relatively large and the arms quite flat. The 
coloration is usually pleasing and often brilliant in life, only a few species being inconspicu- 
ous, or dull. Preserved specimens are usually brown or gray of some shade, or more or less 
bleached of color, and give little indication of their beauty in life. In Linckia, Leiaster, 
and Ophidiaster, at least, autotomy is common and in some species of Linckia at any rate 
asexual reproduction by that means occurs regularly throughout youth, if not after matur- 
ity is reached. 
The basis for generic divisions in the family have long been recognized in the distri- 
bution and arrangement of the papule, with the concomitant arrangement of the plates 
of the abactinal skeleton, the form of the rays, the character of the adambulacral armature, 
the character of the skin, and the nature and form of its granules. Spines, particularly 
erect movable spines, apart, of course, from those of the adambulacral and oral plates, 
rarely occur in the family, and their presence may be of generic importance. Differences 
of opinion in regard to the classification of the family are due to the relative weight given 
to these various characters. My studies have led me to believe that it is best for practical 
purposes to make the primary basis of division in the family the arrangement of the abac- 
tinal skeletal plates. It must be admitted that certain Ophidiasters, and a few species in 
other genera, give some trouble in this particular, especially when the specimen is either 
youthful or senescent, but, on the whole, the arrangement is in general easily determined 
and most of the doubtful cases show their true affinities in some other obvious ways. 
I have felt it desirable to revive all? of the generic and subgeneric names proposed by Gray 
in 1840, but rejected by Perrier in 1875 and by later writers. As Fisher says (in litt.) 
“Gray had a pretty good ‘generic sense’,’”’ and although his diagnoses are poor and often 
worthless, his proposed genera and subgenera usually prove to be natural groups. One of 
them, Cistina, has been rejected because no specimen seems to be extant and the diagnosis 
indicates a most unusual feature in an ophidiasterid, but as the type was said to be from 
the west coast of Colombia, it is quite possible that when that now virtually unknown 
coast is thoroughly explored, this remarkable sea-star will again be found. At any rate 
it does not seem to me we have any right to reject the genus until we do know fully the 
Colombian marine fauna. 
In drawing the line between Linckia and Ophidiaster and its allies, I have not found 
that the character of the adambulacral armature is of much assistance. Of course the 
difference, in this matter between a typical Ophidiaster and a typical Linckia is obvious, 
but in each genus there are species which approach the other, and in such cases I have 
‘While I greatly regret to differ with Fisher on a matter of nomenclature, I see no valid reason for abandoning 
Verrill’s name for this family proposed several years before the commonly used name Linckiid. Other things being 
equal, priority may well fix family names. 
* Except Acalia. See footnote, p. 63. 
