EASTMAN : DESCKirTIONS OF BOLCA FISHES. 31 



Chaetodonts has already been suspected by Smith Woodward. It is probable 

 that they represent types of distinct genera, but for the present they may be 

 most conveniently included within the limits of Pygaeus, as purposely extended 

 by Agassiz. It is evident that some of these forms are closely related to 

 modern Acanthuridae, the chief differences consisting in the great development 

 of the dorsal spines, and the fact that the maxilla and premaxilla are distinct. 

 The latter condition is alone sufficient to warrant the retention of these larger 

 species of Pygaeus amongst the Chaetodontidae, rather tlian amongst the Acan- 

 thuridae, or so-called " Acronuridae " of Giinther, and Teuthidae of Jordan. 

 On the other hand, the teeth are much stouter than in living Chaetodonts. 



We have now to offer the description of a new species of Pygaeus, as con- 

 strued in its broader sense, no division of this genus being at present attempted.^ 

 The type-specimen formerly belonged to the Marchese di Canossa Collection, a 

 part of which was purchased some months ago for the Museum of Comparative 

 Zoology. The choice of a specific title has been determined by the desire to 

 commemorate the labors of the master in this field, his name not being simi- 

 larly associated with any other member of the Bolca fauna. 



Pygaeus agassizii, sp. nov. 



(Plate 2.) 



D. 10 + 9; A. 5-1-8 ; V. 5; P. 17 or 18. 



A comparatively large species, attaining a total length of about 19 cm. 

 Maximum depth of trunk contained twice, and length of head with opercular 

 apparatus three times in the total length to base of caudal fin. Dorsal fin 

 arising immediately 1)ehind the occiput and extending as far as the caudal 

 pedicle with ten subequal spines and nine articulated rays, the latter not 

 longer than the former, and not produced into an acute lobe in front. Anal 

 spines gradually increasing in length and stoutness from the first onward, the 

 fifth e(iualling the foremost articulated ray in length, and longitudinally 

 striated. Articulated portion of the anal corresponding in size and position 

 to the articulated dorsal. Abdominal vertebrae 10, caudal 13. Large incisi- 

 form teeth present in front, gradually diminishing in size posteriorly, apjxar- 

 ently in single series ; maxilla and premaxilla clearly separate. Scales small, 

 those of the posterior part of the body in the form of shagreen-like calcifications 

 and tubercles. Neural spines of abdominal region and all of the interspinous 

 bones much expanded ; pelvic bones strongly developed. No lateral caudal 

 spines. 



1 The writer is indehted to President Jordan, than whom is no Iiigher authority, 

 for the suggestion that " Pi/qneus, and y)ossil)ly Apoatntiis also, should he taken as 

 representing a distinct family, which would occupy a more central ])Osition near 

 the common ancestry of Acanthuridae Chaetodontidae, and Siganus ( Teuthis) " (Ittt., 

 May, 1904). 



