COMPAEATIYE MOEPHOLOGT OF THE GALEODID.E, 389 



beginning from the 3rd, and stretching back the whole length of the body. This 

 condition is still traceable in Galeodes, but in the gre^xt majority of Arachnids it is 

 obliterated by fusion, the terga in front of the Avaist fusing to form a shield. The 

 diflerent forms of the shield, furtlier, require us to assume that it arose in the different 

 families from different arrangements and fusings of the primitive free terga. 



This is incompatible with the large cephalic shield of a Limulus [or the smaller but not less compact 

 shield of the Euryptcrids], which is far too specialized a structure to have given rise to the free terga of 

 Galeodes, or the many different arrangements of fused terga which now form the various cephalic shields 

 of Arachnids. 



Further, the cephalic shield of Limulus and the Euryptcrids, so far as the evidence from Apus and the 

 Trilobites goes, did not arise from fused terga, i. e. if we use the term "tergum" to signify a definite 

 plate-like thickening of the dorsal surface of a segment connected with the terga of the neighbouring 

 segments, and with the " sterna," by means of flexible membranes. 



4. Stei'iia. — There can be little doubt that the ancestral form possessed segmental 

 sternal plates all along the ventral surface, which have been variously modified and 

 obliterated in the different Arachnids (PI. XXIX. fig. li). The first obliteration was 

 doubtless due to the distortion of the 1st segment, whereby the first sternite became the 

 rod-like support of the labium {\). 315), and the 2nd was more or less crushed between 

 the coxEB of the pedipalps. The 3rd either persists but slightly altered (Spiders), or has 

 been variously modified and ol)literated, in Thclyphomis by transverse infoldings, in 

 Scorpio by lateral compression from the coxa? of the legs. The sternite of the genital 

 segment was also, no doubt, early ol)literated, perhaps partly owing to the folding back 

 of the limbs of that segment over the genital aperture and partly to the formation of a 

 waist. 



From these two points of initial obliteration of sternites we have almost all stages of 

 further obliteration, culminating in the Galeodidoe. 



It is quite impossible to deduce a racial form with a regular seines of segmental sterna from an animal 

 like Limulus or Eurypterus, in which the original sternites of the first six segments have been so 

 completely obliterated. In order to develop into a typical Arachnid, Limulus would have again to undo 

 its specialization and once more develop its lost sternites, lost on the ce])halothorax by the extreme back- 

 ward prolongation of the mouth, and, having recovered them, specialize them along the lines followed by the 

 different Arachnids. This, I think, is next to impossible. If the mouth of a Limuloid did travel once 

 more to an anterior position, deep traces (scars, as it were) of its former Limuloid specialization would 

 persist ; it could not return so exactly backward along the path of its development, and once more recover 

 its segmental sternites in their original position, order, and undifferentiated condition as would 

 be necessary for it to develop into an Arachnid. The subsequent modifications of a Limulus or of a 

 Eurvpterid would be the modifications of a Limulus or Euri/pterus, and not of a form showing primitive 

 undifierentiated segments, from which alone, I believe, the Arachnids can be deduced. 



5. Intersegmental MemJjranes. — The presence of terga and sterna implies also the 

 presence of flexible intersegmental membranes. These still persist in various degrees in 

 the Arachnids, although in the anterior region they have been to a great extent folded in, 

 owing to the muscular compression of the cephalothoracic segments. In Galeodes ^exQ-eal 

 such membranes persist between the terga of the cephalothorax. In Scorpio and Chernes 

 the terga have fused to form a cephalic shield, but tlie lateral membranes persist. The 



