Page:The Eurypterida of New York Volume 1.pdf/311

 eyes, which occupy the entire space between the border and a median ridge; (5) this distinct median ridge, also a character not observed in later stages, and finally the abdominal segments which are (6) less in number and (7) lack all differentiation between preabdomen and postabdomen as far as the length of the segments is concerned, all being of equal length.

We may now consider some of these larval characters in somewhat greater detail.

The carapace appears to have had more distinctly the nature of a definite shield than at any later stage, for it is here most clearly set off by a flat thickened border of equal width. The same impression is also conveyed by the rounded postlateral angles, well seen in figures 1 and 3.

The large size of the carapace, its distinction from the abdomen and the broad border, are features which this larva has in common with Limulus [text fig. 24], with the difference, however, that in the latter the border is not continuous over the posterior margin. But there are still other features inviting comparison with the larva of Limulus, notably the median ridge, which produces a trilobation of the carapace directly corresponding to that of Limulus in the so called "trilobite stage." It further appears that the number of segments of the abdomen may be the same as that observed in the larva of Limulus shortly after hatching, that is nine including the rudiment of the caudal spine. Eight (including the telson) are distinctly visible in the original of plate 51, figure 4, but it is probable that the first segment is hidden under the carapace. As in the larva of Limulus these segments are not differentiated, at least in their dorsal view and the telson spine is still extremely rudimentary. But herewith all similarities end.

The uniformly terete form of the abdomen of this larva is in strong contrast to the broad, posteriorly widening abdomen of the Limulus larva. There is little doubt that the latter represents a tachygenetic