242 B. J. L. Guppij — Foraminifera of Trinidad. 



marked at the initial or aboral end of the shell. The next segment 

 invests this almost entirely, leaving only a small part of the initial 

 chamber uncovered, and each succeeding segment invests the portion 

 of the shell already formed in the same manner iintil the last 

 segment is added. This last segment is of a size equal to or larger 

 than the whole of the previous segments together, and does not 

 (as the previously formed ones did) completely invest the preceding 

 segments, but only to the extent of about one-half of the longer 

 diameter or length of the shell. This chamber further differs from 

 the preceding ones in being oval in cross section instead of circular. 

 The mature shell thus described (PI, VIII, Figs. 1-4) recalls to mind 

 Zingulina, which appears to be its nearest relation. From the first 

 the aperture of Gonatosphcsra is an elongate slit, while the section of 

 the shell at right angles to the axis of its growth is circular. It is 

 thei'efore evident that an elongate aperture is not necessarily induced 

 by a flattened or compressed contour of the shell, as in Vaginulina. 



Diagram- Section of Gonatosphcera. 



As to ZiH^nZina, Carpenter says (Introd., p. 164) that it is nothing 

 but a compressed Nodosana whose transverse section is oval, and 

 whose aperture has undergone a corresponding elongation. But the 

 case of Gonatosi)hcera shows that a shell of a circular section may 

 have an elongate slit-like aperture, and that moreover a shell having 

 such a character in its younger stages may develop a subsequent 

 segment of compressed or oval contour, as if indeed the shape of the 

 aperture had ultimately a greater controlling influence than the 

 shape of the shell. I think these facts tend to establish the validity 

 of LinguUna as a generic group near which Gonatosplicera might be 

 placed in the family Nodosaridte. 



As regards the minuter structure of Gonatosphara, the test is 

 minutely and closely tubulated like that of the unornamented 

 NodosaricB, but it seems to be composed of two or more layers of 

 shell-substance. 



Brady (Chall. Eep,, p. 517) regards LinguUna as representing 

 transition stages between Nodosaria and Frondicularia. This may 



