Carpenter — Relationships between Classes of Arthropoda. 3^37 



mentioned is, on the whole, an ascending sequence. The reduction 

 and fusion of segments in the hind-body may be regarded as evidence 

 of specialization, as strong as similar fusion in the head-region. The 

 Scorpions have a well-developed abdomen with twelve free segments. 

 In the Pedipalpi the twelve segments are still recognizable, but the 

 hindmost are reduced and crowded. In the Spiders only a single 

 genus (Liphistius) retains any certain trace of abdominal segmentation ; 

 all other spiders have the abdominal segments fused into a compact 

 hind-body constricted off from the cephalothorax by a narrow waist 

 (the pre-genital segment). In the Phalangida the segmentation of 

 the abdomen may be more or less apparent, though the anterior 

 segments tend to disappear, or to become fused with the cephalothorax. 

 And lastly, in the Mites we find cephalothorax and abdomen fused into 

 a single ovoid mass, all trace of segmentation having vanished. 



Now, it is surely a very striking fact that we find this condensa- 

 tion and fusion of the hind-body region correlated with a replacement 

 of lung-books by tracheal tubes as breathing-organs. In the Scorpions, 

 four pairs of lung-books are present ; in the Pedipalpi and the Avicu- 

 larian Spiders, two pairs, which belong, however, to the genital and 

 post-genital segments, and do not therefore correspond with any of the 

 Scorpions' lungs; in the vast majority of Spiders one pair only, the 

 hinder pair being replaced by tracheal tubes. And in the Phalangids 

 and Mites, no lung-books whatever are present, the breathing being 

 entirely tracheal. Considering more particularly the Spiders, it might 

 seem needless to insist that the Dipneumonous families are higher than 

 the Tetrapneumonous, were it not that the true relationship of the 

 Arachnid orders depends so much on the appreciation of this point. 

 For if the relationship between the two great divisions of the Spiders 

 be as just stated, it is certain that among the Arachnids, lung-books 

 are more primitive organs of respiration than are tracheal tubes. 

 Compare the two pairs of respiratory slits in Avicularia, all leading to 

 lung-books (fig. 4a), with the similarly situated two pairs in one of the 

 lower Dipneumonous Spiders — Dysdera (fig. 4b), for example — in which 

 the hinder pair lead to tracheal tubes ; then with a higher type such 

 as Anyphsena (fig. 4c), in which the hinder pair of openings have 

 coalesced to form a median slit, half-way back along the ventriil 

 surface of the abdomen ; and, lastly, with a highly organised Spider 

 like Epeira (fig. 4d), where the median slit is far back just anterior to 

 the spinners. Is it possible to regard such a series without being 

 forced to the conclusion that the arrangement in Avicularia is the 

 most primitive, in Epeira the most specialized ? And a corresponding 



2E2 



