retained the last long leg and several of the anterior limbs scattered about the carapace, furnished the evidence for the two pairs of short limbs in his restoration. Woodward figured in front of these two feelerlike appendages which in Beecher's restoration of S. excelsior are replaced by a first pair of short walking legs, the latter together with a chelicera having been successfully prepared by Clarke [1888, pl. 26A].
An example from the dark shales at Otisville has shown by preparation a series of four legs on one side and this species exhibits in the formation of its carapace the typical characters of a Stylonurus. The structure of this species, S. cestrotus, has suggested a different conception of the limbs of Stylonurus, for, (1) the last two legs are of distinctly different length, (2) the second and third pairs of legs are so long as to form a progressively growing series with the fourth and fifth pairs. The question then arises whether Stylonurus had been incorrectly understood or whether the Otisville form represents a new and different group.
A survey of the 13 species, cited below as falling under the head of Stylonurus by virtue of the characters of their carapaces and abdomina, shows that, besides the original representatives of the genus, S. powriei and logani, of only four species are specimens known that retain sufficient fragments of the legs to indicate their structure. These are S. macrophthalmus and S. elegans Laurie, S. cestrotus Clarke and S. scoticus Woodward. Taken together these demonstrate two important facts; (1) that the legs increased in length quite regularly backward instead of being divided sharply into two different sets as represented by Woodward and Beecher: an anterior one of very short legs and a posterior one of exceedingly long ones, (2) that there are three distinct types of legs among species referred to this genus.
As already stated, Woodward seems to have combined the aspects of his specimens of S. powriei and S. logani to effect his restoration of S. logani; the former furnishing the two long legs, the other the preceding two short pairs. In considering the question of the two "subequal" long pairs of swimming legs, it may be observed that the
