PARKER: MALDIVE CEPHALOCHORDATES. 41 



The specimen obtained by ^Ir. Agassiz is exceptionally well preserved, and 

 the following notes based upon an examination of it are intended to supple- 

 ment former descriptions. 



The length of the specimen is 9 mm. and its depth 0.8 mm., being slightly 

 smaller than the " Challenger" specimen, though having almost exactly the 

 same proportions. Cooper's specimen had the considerable length of 21 mm., 

 though, as his figure shows, its depth was proportionally as great as that of the 

 " Challenger " specimen. Tattersall's largest specimen was 8.5 mm. long and 

 1 mm. deep, thus agreeing fairly well with the measurements of the small 

 individuals already given. Tattersall's drawing (1903t>, Fig. 16), however, 

 represents, probably by mistake, a more slender animal. Notwithstanding the 

 slight differences in the dimensions of the various specimens, thej^ probably all 

 belong to one species. 



The fins in the specimen taken by Mr. Agassiz were in perfect preservation. 

 The caudal fin (Plate 1, Fig. 2) is oval in outline and almost exactly sym- 

 metrical dorsoventrally. The chorda extends through its major axis, and ends 

 close to its posterior edge. The fin dorsally and ventrally is marked with 

 numerous delicate radiating striae. 



Dorsally the caudal is continuous with a dorsal fin, which can be followed to 

 the anterior end of the animal. Near the tail (Fig. 2) it is moderately high, 

 but it gradually becomes lower and lower, till near the anterior end (Fig. 1) 

 it is about one-third as high as at its posterior limit. In the posterior region 

 a row of low fin-ray chambers can be seen (Fig. 2). These increase in height, 

 anil near the middle of the trunk reach the outer edge of the fin. At the 

 anterior end they rapidly diminish in height, and cease near, the eye-spot 

 (Fig. 1). They contain no fin-rays. 



A low ventral fin without fin-ray chambers connects the caudal fin with the 

 right nietapleural fold. 



Anteriorly the dorsal fin is continuous with a ro.stral fin (Fig. 1) which ex- 

 pands ventrally and posteriorly to join the buccal hood on the left side. 



The number of myotomes in the specimen under examination is sixty- 

 seven, corresponding in this respect exactly with the " Challenger " specimen. 

 Cooper's specimen contained at least sixty, and Tattersall's sixty-five, though, 

 as the authors state, neither of these counts can be relied upon as accurate be- 

 cause of the poor .state of the material. 



The myotome formula for this species had been provisionally stated by Kir- 

 kaldy (1895, p. 320) as 36 + ] 6 -f- 15 ~ 67. This was based upon a tentative 

 statement by Giinther (1889, p. 43), in whose specimen the atriopore could 

 not be identified with certainty. In our specimen the atriopore and anus were 

 di.?tinctly visible, and the myotome formula proved to be 46 -|- 10 -j- 11 =67. 



The notochord, which is well developed, reaches from very near the anterior 

 edge of the rostral fin almost to the posterior limit of the caudal. At both 

 ends it projects well beyond the myotomes. 



The nerve tube contains in its anterior end a well developed eye-spot and. 



