JuiT 5, 1907] 



SCIENCE 



13 



from a newcomer in the field, but it is easily 

 the most important publication on starfishes 

 which has appeared in America, at least since 

 Alexander Agassiz's " North American Star- 

 fishes." Although Dr. Fisher is chiefly con- 

 cerned with a systematic account of the 

 Hawaiian Asteroidea, particularly those col- 

 lected by the Albatross in 1902, many im- 

 portant morphological points are discussed and 

 the geographical distribution of the species is 

 given careful consideration. The introduction 

 states that the Albatross collection consists of 

 about 1,650 specimens, representing 60 species, 

 of which the author considers 52 are new to 

 science. This surprisingly large percentage 

 of novelties is not so remarkable, however, 

 when the situation of the Hawaiian group is 

 considered, and it is remembered that, previ- 

 ous to this report, only eleven species (and at 

 least two of these erroneously) were recorded 

 from the islands. The classification used is a 

 compromise between the two fundamentally 

 different systems of Perrier and Sladen, with 

 due reference to the work of Verrill, but un- 

 fortunately Fisher does not explain what prin- 

 ciples have governed him in his choices, and 

 we are left in the dark as to what characters 

 he considers of the greatest importance in 

 determining relationships. He apparently 

 gives considerable weight to the absence, or 

 presence and form, of paxillse, superambu- 

 lacral plates and pedicellarise, all of which, 

 while of value in classification, reveal notable 

 diversities in nearly related - forms. On the 

 whole, however, there can be little question 

 that Fisher has shown excellent judgment and 

 his classification is probably as satisfactory 

 as can be devised in the present state of our 

 knowledge. The nomenclature used is ad- 

 mirable, for, following the sensible lead of 

 Bell and Verrill, adherence is given to the 

 International Code, and the pre-Linnasan 

 names, resurrected or maintained by Sladen 

 in his great Challenger monograph, are re- 

 jected. While this course results in the alter- 

 ation of some familiar names (such as the 



from U. S. Fish Commission Bulletin for 1903, 

 Part III., pp. 987-1130, pis. I.-XLIX. Issued 

 June 30, 1906. Washington: Government Print- 

 ing OflBce, 1906. 



change of Crihrella to Senricia and Pahnipes 

 to Anseropoda) , there can be no doubt that it 

 is a long step towards a stable nomenclature 

 for starfishes. Following the introduction, 

 ten pages are given to a synopsis of the 

 Hawaiian Asteroid fauna and a discussion of 

 its relationships. The most interesting points 

 here brought out are that, although starfishes 

 were taken at 126 stations, the Albatross col- 

 lection consists almost entirely of sublittoral 

 and continental forms, only three or four 

 species having been taken in less than 20 

 fathoms and fewer than ten in more than 500 

 fathoms ; that only one species previously 

 recorded from the islands was taken; that the 

 Hawaiian starfishes are distinctly Indo-Pacifie 

 in their relationships rather than American, 

 notwithstanding that the ocean currents pass- 

 ing the islands are west-bound; that these 

 west-bound currents have only brought five 

 species of starfish to the islands ; and that 

 seven or eight Hawaiian species find their 

 nearest relatives in the Atlantic Ocean or 

 Caribbean Sea. The following 116 pages are 

 taken up with the account of the Hawaiian 

 starfish fauna, and open with an artificial 

 key to the nineteen families recognized as 

 occurring in the area under consideration. 

 This key is clear and well arranged, and while 

 obviously artificial in a few particulars (as 

 such keys are of necessity bound to be), it 

 will doubtless prove more useful than any 

 similar key hitherto published. It is prob- 

 able, however, that further investigation will 

 reduce the number of families recognized by 

 Fisher, for they are certainly of very unequal 

 value as at present arranged. Such families 

 as Mithrodiidfe and Heliasteridse are certainly 

 much less satisfactory than Astropectinidaj 

 and Linckiidse. Following the key, the fami- 

 lies are taken up in their natural sequence, 

 the synonymy being followed by an artificial 

 key to the genera, and each genus, when con- 

 taining more than a single form, is introduced 

 by a key to the species. These keys are ad- 

 mirable in matter and form and make up in 

 a measure for the absence of any diagnoses 

 of families and genera (except those which 

 are new). The descriptions of the new 

 genera and species, and the remarks on 



