REVISION OF THE KEYHOLE UECHINS^ — CLARK 441 



are 49-52 by 57-62 mm. From Bahia, Brazil, there are two bare half- 

 grown tests, one of which is in fragments. From Itabapuana, Brazil, 

 there are nine bare tests of young individuals, 37 by 42 mm. up to 59 

 by 68 mm., the smallest with the apex very high (6 mm.) and far back 

 (16 mm.), and the posterior lunules I and IV are still open at the 

 margin. From off Barbados in 100 fathoms there is a very young 

 individual collected by The Hassler, 12 by 13 mm. with the paired 

 lunules indicated only by slight notches in the test margin while 

 the oral ambulacral areas II and IV are little more than 4 nun. long, 

 and their distal width is nearly 3 mm. 



In the lot of Mellitas from the United States National Museum 

 tliere are seven specimens to be referred to this species, and though 

 all are young they are quite typical and are to be considered para- 

 types. The largest is a bare and water-worn test 57 mm. long by 

 65 mm. wide, v^ith the apex 6 mm. high and 23 mm. back of the 

 margin. It is said to be from the coast of the State of Sao Paulo, 

 Brazil. From Itajahy, Brazil, are five young specimens, three of 

 which are notable for having retained most of their spines; the 

 general color is a light greenish gray, but the oral ambulacral areas 

 are more or less violet -brown in rather noticeable contrast ; areas II 

 and IV are conspicuously wide, 22 by 13 mm. in the largest specimen, 

 which is 48 by 65 by 4.5 mm.; the smallest is only 42 by 45 mm. 

 The bare tests are also small individuals, and the larger is notable 

 for its pale blue-green color. From the beach at Santos, Brazil, is 

 a young bare test 43 by 40 mm., with the oral ambulacral areas II 

 and IV scarcely 20 mm. long though fully 11 mm. wide. 



Even extreme examples of this species do not equal the wide pro- 

 portions of lata^ but they do exceed the widest quinquiesperforata 

 available for comparison. In this particular it is probable that no 

 hard and fast specific lines can be drawn, but in the great width 

 of the oral ambulacral intermediate areas, especially in II and IV, 

 a very trustv»'orthy specific character is found. Of course, in young 

 individuals (less than 40 m.m. long) it is not so marked as in adults. 

 The great thickness of the anterior half of the test as compared 

 with the posterior is also a useful character, and in mature speci- 

 mens the great length of the unpaired lunule is a notable feature. 



This seems to be the characteristic MeUita of Trinidad and the 

 eastern Venezuelan coast. How far the range extends to the west 

 is still to be determined, but southward it seems to include prac- 

 tically the whole coast of Brazil, even to the State of Santa 

 Catherina. The puzzling thing about this vast range is that quin- 

 quies perforata also seems to occur on the Brazilian coast; at least 

 there are bare tests in the Museum of Comparative Zoology taken 

 by the Thayer expedition at Maranhao that are undoubtedly the 



