524 NATURAL SCIENCE. sept 



palaeontology will ever throw mucli light on the ancestry of the two 

 classes ; it is to the structure and development of living forms that we 

 must look for our information. 



In the embryo of Trochosa the limbs, which afterwards become so 

 differentiated as the chelicerae, and also the rudimentary limbs on the 

 abdominal segments, which afterwards disappear altogether, show 

 indications of jointing like the legs ; this is strong evidence of the 

 former ambulatory function of all the limbs. Another observation of 

 great interest made by Jaworowski is that in the cephalothoracic 

 appendages of the embryo Trochosa (except the chelicerae) there are 

 parts which correspond to the coxopodite, endopodite, exopodite, and 

 epipodite of a crustacean limb. The epipodite is more obscure in 

 the hinder segments, and, in all cases, disappears. From the fact 

 that the epipodite of a crustacean often bears a gill, and that the 

 spider's epipodite is a transitory embryonic structure, Jaworowski 

 infers the descent of the air-breathing Arachnids from water-breathing 

 ancestors. The epipodite of Trochosa is only mentioned in his most 

 recent paper (3). It is rather amusing that Simroth in his book, " Die 

 Entstehung der Landtiere," should have used Jaworowski's first 

 paper (2) with the description of the endo- and exopodite of Trochosa 

 in support of the exactly opposite conclusion : that water-breathing 

 crustaceans have branched off from ancestors that lived on the land ! 



The aquatic representati'v'e of the Arachnids, however, is the 

 King Crab [Limulus), whose gills, borne on the hinder faces of tlie 

 flattened abdominal limbs, have been repeatedly compared with 

 the lung-books of spiders and scorpions. Lankestt r (5) and Kingsley 

 (6) suggested tlie formation of the arachnid lung-book by the invagina- 

 tion of an appendage intermediate in form between the gill of 

 Limuhis and the pecten of a scorpion. MacLeod (7) proposed that 

 the flattened appendage of Limultts might become fused by its edges 

 with the body-wall, so as to enclose the gill-lamellae. Laurie in a 

 recent paper (8) inclines to the latter view. He finds that in the 

 embryo of Scorpio fulvipes the lamellae lie horizontal, and parallel to 

 the ventral surface, directed forwardly and outwardly from the stig- 

 matic opening. Such an arrangement would require the invagination 

 of an appendage bearing the lamellae on its front face, whereas the 

 appendages of Limulus bear them on the hinder face, and so Lan- 

 kester's theory would require the complete reversal of the gills. No 

 such difficulty is raised by MacLeod's view. It is noteworthy that 

 in the adult Scorpio the lamellae take up a different position from 

 what they have in the embryo, lying vertically and parallel to the axis 

 of the body. In the adult spider they retain the horizontal position, 

 so that in this respect the spiders show the more primitive arrange- 

 ment. 



In most Arachnids a pair of coxal glands is found near the base 

 of one or more of the pairs of legs; these structures, which open to 

 the exterior only in the embryo and become ductless in the adult, 



