282 NEW YORK STATE MUSEUM 
A further question relating to the legs, suggested by the Otisville 
specimen of S. cestrotus, is whether the second and third pairs 
were as short and as nearly equal in size as represented by Woodward 
and Beecher. Here again all the types which retain these legs; those 
of S. cestrotus, 5. macrophthalmus, S. elegans and 
S. ornatus, give direct evidence that they also formed a posteriorly 
increasing series, the last being the longest and that, on the whole, they ' 
were much longer than represented in the restorations. The specimen 
of S. logani [text fig. 63] which served as the basis of the restorations 
exhibits a fragment of two segments of one of these legs attached to the 
body and is surrounded by two (three in first figure, 1864) detached anterior 
legs. The smaller of these undoubtedly represents the first pair of postoral 
limbs, as clearly indicated by its rapid contraction and the shortness of 
the segments. As the first pair Woodward introduced antennae and con- 
stituted as the second pair what we believe to be obviously the first, while 
the other, which has about twice the length of the former, to us repre- 
sents the second pair. The third pair is only represented by the small 
fragment still attached. Itis quite clear that there is considerable differ- 
ence in length between the first and second legs and a still greater difference 
between the second and third legs. The relative lengths of the first and 
second legs are well shown in Laurie’s type of 5. macrophthalmus 
[text fig. 64] and the relative great length of the third pair is evident 
in S. elegans [text fig. 62] where it nearly equals the fourth pair 
and surpasses it in width. Looking back again, in relation to these legs, 
to Drepanopterus, the ancestor of Stylonurus, we find that there the 
first three pairs also form a series increasing in length backward. 
The combined evidence, then, of the material at present available 
is that the legs formed an approximately continuous series, the anterior 
members being longer than hitherto supposed and the last two not so greatly 
surpassing the others in length and not of so nearly equal length. On the 
basis of these facts we have drawn a new reconstruction of 5. excel- 
sior, although only the first pair of legs of that species are known, using 
_ mainly our specimen of S. cestrotus for this reconstruction. _ 
