132 PROFESSOR E. RAY LANK ESTER. 



larly on its dorsal surface the alimentary tract is enclosed in a 

 gastric canal (g c, in PL VI, fig. 5), and the dorsal artery or 

 aorta in an arterial canal (a c). In the Scorpions, also, the 

 connective tissue continued from the sides of the median plate 

 is tough and dense, so as to form an expansion of the entoster- 

 nite in the form of a right and left "posterior flap'^ {p f, in 

 ^late VI, figs. 5, 6), by which the entosternite is brought into 

 continuity with the lateral walls of the body. Thus, in the 

 Scorpions the entosternite assumes the form of an obliquely- 

 placed "diaphragm^' — a name applied to it by Newport — 

 cutting off the cavity of the prosoma almost completely from 

 the cavity of the mesosoma. This lateral extension of the en- 

 tosternite does not occur in either Limulus or Mygale. 



The various processes and ridges of the entosternites of 

 Limulus and of Scorpio are named in the description of PI. VI. 

 In attempting to determine the equivalent structures in the 

 entosternites of these two animals it is necessary to make use 

 of the indications afforded by the attachment of the muscles. 

 This is a subject which I do not propose to treat here, since I 

 have dealt with it in the paper already mentioned in the ' Trans- 

 actions of the Zoological Society.^ I may, however, briefly 

 point out that only that small portion of the Scorpion^s ento- 

 sternite which immediately roofs in the neural canal appears 

 to be equivalent to the ''body^^ of the entosternite of Limu- 

 lus. The right and left anterior processes (rap, lap, in the 

 figures) appear to be equivalent in the two animals, whilst the 

 two pairs of lateral rod-like tendons in the Limulus entoster- 

 nite (alr, plr) are represented by the latero- median processes 

 {Im p) of the Scorpion. There is nothing in Limulus cor- 

 responding to the subneural arch [sup) of Scorpio, nor to its 

 anterior inferior pair of delicate tendons (asp). 



It is interesting to find in Mygale a much closer agreement 

 with Limulus as to the form of the entosternite than we can 

 detect in Scorpio. The two pairs of lateral rod-like tendons 

 are represented by three similar pairs in Mygale (PI. VI, fig. 4) ; 

 the bifurcated posterior median process is present in both, and 

 the general form of the "body^' with its neural fossa (n f) and 



