184 



America, and is more or less variable in its characteristics, but 

 apparently never produces shells so large, thick, or flat-sided as 

 those from Cuba. In fact, these features serve to differentiate 

 Arctostrea atkinsi from nearly all allied forms. Woods (1913, 

 p. 342-355) has illustrated, under the name Ostrea diluviana, 

 several shells belonging to Arctostrea. Among them are some old 

 individuals showing considerable thickening of the valves, as in 

 his fig. 127, p. 353, fig. 99, p. 348, or fig. 138, p. 354. None of 

 these, however, shows the great size or flat sides of A. atkinsi. 

 Coquand (1869, p. 76, pi. 29, fig. 1, 2, labelled Ostrea colubrina) 

 figured as an example of Arctostrea pectinata (Lamarck) a very 

 large oyster which measures 275 mm. along the curvature of the 

 anterior margin. This specimen is of about the same size as the 

 one from Soledad, and agrees further in having the muscle scar 

 almost marginal. A. pectinata has, however, coarser ribs than 

 A. atkinsi, and the sides are not flattened. The only species 

 which I have been able to find which shows this latter character- 

 istic is Arctostrea rectangularis (Roemer), as figured by Coquand 

 (1869, p. 187, pi. 72, fig. 5-11). Roemer's (1839, pi. 18, fig. 15) 

 original figure merely suggests a lateral flattening, but that pre- 

 sented by Coquand shows a shell in many respects like A. atkinsi. 

 The distal portion of the right valve is flattened and nearly 

 smooth along the middle, and the interlocking projections along 

 the margins of the valve are of about the same size as in the shell 

 from Cuba. A . rectangularis is, however, a smaller shell, and be- 

 comes much narrower distally. 



Horizon and locality. — Fragments of three right valves of this 

 species were obtained about 8 feet below the surface, during the 

 digging of a well on the Santa Rosalia property near the Harvard 

 Botanical Gardens on the estate of Mr. Edwin F. Atkins, south- 

 east of Cienfuegos, Cuba. The exact horizon remains unknown, 

 but is doubtless within some subdivision of the Cretaceous. The 

 nearest ally, Arctostrea rectangularis, is found in the Neocomian, 

 and A. carinata is characteristic of the Cenomanian, so that the 

 upper part of the Lower, or the base of the Middle Cretaceous, is 

 suggested. Similar oysters, such as A . pristiphora (Coquand) and 

 A. pectinata (Lamarck) are, however, found in the Campanian and 

 Santonian, so that it is not impossible that our species is really 

 from the Upper Cretaceous. The few fossils found with the 

 oysters have been examined by Dr. T. W. Stanton, who reports 

 the presence of a very oblique and somewhat peculiar type of 

 Cucullaea, and an Echinoid which appears to be a Holectypus, a 

 genus which, according to Doctor Stanton, ranges from the 

 Jurassic through the Comanchean. Unfortunately these two un- 

 described species have no great stratigraphic significance. 



G. F: Matthew (1875, p. 29-30) long ago announced the pres- 

 ence of Cretaceous strata along the Damuji, a small river enter- 

 ing the western end of Xagua Bay. He stated that the limestone 

 at Constancia Landing and on the Concepcion estate, about 10 

 miles northwest of Cienfuegos, contains several species of Hip- 

 purites, Caprinella, and Caprotina, also corals, Oliva, Conus, an 



