990 
BULLETIN OP THE UNITED STATES FISH COMMISSION. 
meant that the family holds an intermediate position, because nothing 1 could be much 
farther from the truth. But since the Luidiidse are a very specialized offshoot from 
the proastropectinoid stock, there is no other position for the group. Similar apolo- 
gies should be. made for the Benthopectinidse and the Archasteridse. 
In the matter of the orders I have made a compromise between Sladen’s and 
Perrier’s and Yerrill’s modifications of these. Sladen’s Phanerozonia is used in 
nearly the same sense in which he employed the term in the “Challenger Asteroidea.” 
The Linckiidse, being phanerozoniate, are taken out of Sladen’s Cryptozonia and 
placed in the Phanerozonia, in which order they undoubtedly belong ; while the 
Asterinidae (and Anseropodidae) being cryptozoniate, are removed from the 
Phanerozonia. I believe Perrier’s arrangement of the cryptozoniate families under 
two orders (he has apparently abandoned the Yelata) is rather more felicitous than 
that of Sladen under one large heterogeneous order (Cryptozonia), and I have con- 
sequently followed Perrier with small modifications. It must be remembered, how- 
ever, that there is a great difference of opinion concerning the extent of the orders 
of Asteroidea. Some writers follow Sladen and some Perrier. As pointed out 
recently by Professor Verrill (1899), Perrier’s Paxillosaand Yalvata (= Phanerozonia 
emended) are decidedly artificial groups. I do not go so far as Professor Yerrill in 
considering them suborders, because unless some more fortunate grouping of families 
is devised, the difficulty is merely temporarily pigeon-holed. 
Returning to the families, it will be noted that I have followed Yerrill in dismem- 
bering the great “family” of the Archasteridse. I have accepted most of Yerrill’s 
modifications and have raised the Pseudarchasterinae to family rank. I can not agree 
with Professor Verrill that the Pontasterinae constitute a subfamily of the Pluton- 
asteridae, but have relegated the group to the Benthopectinidae (=Pararchasterinae 
Sladen), a position more nearly in accord with the views of Sladen. The sequence 
of families is new so far as the Phanerozonia are concerned. The order starts with 
the more specialized paxillose forms, with very well developed superambulacral ossi- 
cles, and proceeds through those in which the paxillae are often rudimentary or 
reduced and the superambulacral plates not always present to those in which the lat- 
ter are always absent. Then, beginning with the Mediasterinae, the transition 
is gradual toward forms with less paxilliform plates, to those with granular or smooth 
abactinal plates, and bivalved or foraminate pedicellariae; then to the skin-covered 
forms, with stellate or stellate-reticulate abactinal skeleton and low bivalved pedicel- 
lariae. The Linckiidse, though placed at the end, are in many respects distinctly 
related to some of the Goniasteridae. They often have excavate pedicellariae. The 
Gymnasteriidae are rather intermediate forms. In the Spinulosa and Forcipulata the 
sequence is practically that given by Perrier. The Velata are merged with the 
Spinulosa. In many respects the classification of the Asteroidea is difficult, especially 
on account of the number of more or less perfect transitional forms which render 
family boundaries extremely unstable. 
So far as I am aware there have been no special papers on Hawaiian starfishes. 
The few references are widely scattered and a number of records are given second 
hand, the original citation being unknown to me. A number of records are very 
incidental in mention, and it is probable that a few have escaped notice. A bibli- 
ography of papers quoted in the following report is appended. 
