COMPAEATIVE MOEPHOLOGT OF THE GALEODIDiE. 329 



are the stigmatic combs. A less specialized group of setse appeal* iu nearly the same spot 

 on segment 5 of the same figure, and singular modifications, presumably of these rows of 

 setai, occur iu the specimen of G. intrepidus on the 5th, 6th, and 7th segments, also 

 figured (PI. XXIX. fig. 8). 



Lastly, the hypothetical restitution of the abdominal limbs at an earlier stage in their 

 degeneration, shown in PL XXIX. fig. 11 (which will be discussed further in the section 

 on the tracheae), suggests that the sclerite su.rrounding the anus may also be a fused 

 pair of limb-vestiges. 



Abdominal limbs in Arachnids are well known in the mammillse of the Spiders, iu the pectines of 

 the Scorpions, and in the "penis" of Phryims (77). Further, from the presence of the lung-books, 

 vestiges of limbs can be safely recognized in the areas covei'iug these breathing-organs. 



Until a comparatively recent time, the traces of abdominal limbs were thought to be confined entirely 

 to the first six abdominal segments ; in Scorpio we have the genital opercula, the pectines, and four pairs 

 of lung-books. But I claim to have found clear traces of the former presence of tracheal invaginations, 

 and therefore of limbs, on all the abdominal segments of certain Cheruetids (10) and on eight segments 

 of Thebjphonus (15). 



These facts fully agree with what we have just seen of Galeodes, where fairly clear vestiges of at least 

 nine pairs occur. 



While, then, Scorpio * and the Spiders show traces of only six pairs of limbs having been 

 present on the abdomen, three Arachnids, belonging to very different families, bear testimony to more 

 than six. Thelyphonus shows traces of 8 pairs, Galeodes of 9, and Chernes of 10. We may therefore 

 conclude, what indeed we might have safely concluded on almost any theory as to the origin of the 

 Arachnids except one, that the primitive form had limbs on all the abdominal segments. The only 

 theory according to which the abdominal limbs could not have existed in the posterior segments will be 

 discussed in detail in the last section. 



We have further some evidence as to the original character of these limbs, which are now, in the 

 Galeodidse, reduced entirely to plate-like structures. In some Phrynidae the first pair still persists in the 

 so-called "penis" (77, and PI. XXIX. fig. 13) as a pair of jointed limbs folded together iu the middle line 

 and illustrating the formation of the genital operculum. In the Phalangid;e, the long penis or ovipositor 

 is almost certainly a specialization of a pair of filamentous appendages. In Scorpio, the limbs of the 

 2ud segment, the pectines, are filamentous and still bear traces of three joints ; we learn from this that 

 the plates covering the stigmata were once filamentous and jointed limbs. This gives us a series of such 

 filamentous appendages as far as the 6th segment in Scorpio and as far as the 8th in Thelyphonus. That 

 such filamentous limbs did exist on the 4th and 5th segments we have direct evidence, viz. their 

 persistence as mammillse iu the Spiders, which are especially leg-like in the Aviculariidse. 



As a rule, then, the vestiges of the abdominal limbs in Arachnids are now little more than plates or 

 folds, covering the genital and stigmatic apertures, or else are flattened down as hardly distinguishable 

 portions of the ventral segmental sclerites. In a few cases, however, the original leg-like character has 

 not been completely obliterated, these cases occurring in very distinct families and on diflerent seo-ments 

 always in adaptation to the specialization which characterizes these families. 



* I have often thought that the tail-sting of Scorpio was a composite structure, representing a pair of Ics (of 

 the anal segment) fused posteriorly over the anus. Limbs meeting over the anus and pointing backward 

 are found in Scolopendra. Such an origin would explain the remarkable fact that the poison-glands are oaired. 

 Of. further, the anal plates of Galeodes, and the diagrams, PI. XXIX. figs. 8-12, 14. 



SECOKD SERIES. — ZOOLOGY, VOL. VI. 44 



