NO. 1659. AMERICAX SPECIES OF SYNALPHEUS—COUTIERE. 3 



S. longiearpus a belongs in realit}^ to the group Avhich I call the 

 L^KviMANUs group, from the name of the Mediterranean species of 

 Heller, a group which is characterized essentially by a brush of long 

 stiff hairs on the movable finger of the small claw. I have been 

 obliged to recognize eighteen species and subspecies belonging to it 

 upon the American coast, so that the old specific limits have become 

 greatly narrowed; I have retained the name lonr/irarpus for that 

 species which appears to be among the most widelj^ distributed in the 

 region of the Gulf of Mexico and the Bahamas, and which conforms 

 to one of Herrick's types; the species has small eggs and its larva3 

 are zoese; in regard to /S. longicarpus /?, the specimens which I had 

 at first grouped under that name are found to be referable to three 

 very distinct species, each provided with numerous subspecies; that 

 species which corresponds to the type specimen of Herrick is S. 

 peetiniger^ new species; the other two have received the names S. 

 hrooksi and S. herricki. All three have eggs of large size, from which 

 spring my sis larvae, and one of the species must certainly be Herrick's 

 nominal species Alpheus precox^ without its being possible to definitely 

 determine which. 



Save for the preceding exceptions, all the forms, perhaps thirty 

 species and varieties, have had to receive new names. It is a con- 

 siderable number, and surprised me at first. Although several pre- 

 sent very strong resemblances to other forms of the eastern Atlantic, 

 the Indian Ocean, and the Pacific islands, all are peculiar to the 

 American coasts. This is true also of S. lockhigtoni^ which is repre- 

 sented in the Indian Ocean by some closely allied species, which, in 

 turn, are difficult to separate from specimens from the Red Sea, the 

 Mascarene Islands, and from the west coast of Africa. It would seem 

 that these specimens represent local races of a cosmopolitan species. 

 There is, however, a remarkable exception in S. latastei, Chilian speci- 

 mens of which can not be distinguished from Australian. 



CLASSIFICATION OF THE SPECIES. 



In view of the grooving number of species of the genus Synalpheus^ 

 one is led to distinguish among them several groups composed of the 

 more closely allied forms, which may be differentiated in the following 

 manner : 



KEY TO THE SPECIFIC GROUPS OF THE GENUS SYNALPHEUS. 



a\ Supraorbital spines insignificant eoini)ared to the rostrum ; antennules 

 shorter than the antenutt; spines of the basicerite almost equal, the exter- 

 nal always smaller than the stylocerite; external maxillipeds oval, feebly 

 spinous distally ; first segment of the carpus of the second pair of feet 

 very long; following feet cylindrical; ventral hook of the dactyl obsolete; 

 telson with an oval median lobe Comatularum group. 



a^ Supraorbital spines at least equal to the rostrum in importance; antennules 

 at least equal to the antennae; spines of the basicerite unequal, the exter- 



