43 



iH and a width of i 3-16 inches, the dividing ridge in the posterior 

 half ending in a short spine, and the serrated margins, like the one 

 described in the last number of this Bulletin. Fragments of eight 

 abdominal segments which are shown on the stone, have a length of 

 4H inches. The whole has a remarkable position turning almost 

 upon itself. Neither part has any surface markings. 



The term of flat seems more appropriate for the tails of this 

 species, as they represent only very thin impressions, although both 

 sides are present, and thicken only a little where their anterior por- 

 tion connects with the abdomen. 



4. Pterygotus quadraticaudatus, n. sp. Plate III., Fig. 1.) 



One tail found. It resembles in form an irregular square with 

 the anterior and posterior margins of almost equal length, running 

 parallel; the lateral margins, extending at right angles to the anterior 

 one for about one-fourth of their length, then assume a convex shape, 

 and unite at an obtuse angle with the posterior margin. The lateral 

 margin is serrated at its posterior half. Two slightly convex curves 

 of equal length which form the posterior margin, connect in a con- 

 cave indentation at its middle, and from here a faint dividing line 

 originates which is visible for about one-quarter of the length of 

 the tail. 



The surface of the anterior portion is covered with small scale- 

 like processes distinct from each other. The markings on the rest 

 of the surface are obliterated. 



Length of specimen, 2 7 A inches; greatest width, 3^2 inches; nar- 

 rowest part, 2H inches. Found in the water-lime group near 

 Buffalo. 



We have thus, up to the present time, three distinct, specific 

 characters in the tails of the American Pterygotis. We are justified 

 in rejecting the assumption that they may simply represent different 

 stages of development, as in the five specimens of P. acuticaudatus 

 which our museum possesses; the tails range from 1% to 6X inches 

 in length, and all have the same shape without any variation what- 

 ever. On the caudal appendages of P. globicaudatus and P. quadra- 

 ticaudatus, aside from their shape, the surface markings, unless alto- 

 gether obliterated, are so different, that no difficulty can be encoun- 

 tered in identifying them even from small fragments. 



