COMPARATIVE MOEPHOLOGT OF THE GALEODID.E. 383 



Revietv of the Ghtnds in the Arachnids. — The Galeodidse appear to differ from all other large Arachnids 

 in possessing no other couspicuous glands, such as poison- or spinning-glands, although the former were 

 at one time claimed for Galeodes. Their well-developed coxal and INIalpighian glands appear to suffice 

 to remove waste products. In addition to these must be added the glands which open within the 

 genital aperture and which may secrete some glutinous substance (PI. XXXIII. fig. \,gg). 



In Phrijmis there are well-developed coxal glands and ]\Ialpighian tubules. To these, however, must 

 be added spinning-glands, which appear to correspond with the glands opening within the genital 

 aperture in Galeudvs. They belong to the limbs forming the genital operculum, and yield the coarse 

 silk on the egg-cocoon (PI. XXIX. figs. 11, 13,ff(j). 



The Pseudoscorpions have coxal glands, but no Malpighian vessels. On the other hand, they have 

 enormous spinning-glands, opening at the tips of the movable digits of the chelicerie, as discovered by 

 Croneberg, and cement-glands opening behind the genital aperture. These glands appear to develop 

 periodically and probably not concurrently, and it may be safely assumed tliat they utilize a large 

 proportion of the waste jjroducts. 



The silk-spinning powers thus feebly developed in Phrynus, and more strongly in the Pseudoscorpions, 

 Iiave reached their most marvellous development in the Spiders, where, in many cases, a perpetual 

 flow of silk seems to suffice to utilize all the waste products, so that the coxal glands can degenerate, 

 and the Malpighian vessels change their functions. From this point of view, the web-spinning of the 

 Spiders for the netting of prey is but a development of the cocoon-spinning of the Pseudoscorpions. 



In addition to the spinning-glands, the Spiders have well-developed ])oison-glands. 



In Scorjno, in addition to the coxal glands and IMalpighian vessels, there is a pair of large poison- 

 glands in the 'sting' on the anal segment. And in Thelyphunus also, in addition to the well-develoi^ed 

 coxal glands and Malpighian vessels, there is a large pair of glands (stink-glands) which open on the 

 soft membrane round the anus, on each side of and slightly above the anal aperture. It is highly 

 probable that these poison- and stink-glands, like the spinning-glands, utilize waste products. The 

 stink-gland of Thehjphonus is probably homologous with the poison-gland of Scorpio. That these three 

 glands, poison-, spinning-, and stink-glands, are the common derivatives of the setiparous glands is, if 

 not universally, at least widely accepted. 



One of the more striking instances of tufts of hair in one Arachnid occupying the exact position of 

 spinning- and poison-glands in others is to be seen in the hairs on the dorsal surface of the tip of the 

 movable digit in the chelicerie of Phrynus (PI. XXIX. fig. 2, set) . These stout hairs seem deeply set in 

 pits. In the Pseudoscorpions, at exactly the same sjiot, are a group of minute apertures for spinning- 

 glands, and in the Spiders there is the aperture of the poison-gland. Lastly, in Galeodes these setje 

 have vanished, while the setal pores appear to remain open, showing no specialization into glands of any 

 kind, but, as I have suggested, possibly exuding matter which is highly poisonous. 



If the tracliete are added to the spinning-, cement-, and poison-glands as homologous structures, we find 

 that there must have been two series on each side. In the Pseudoscorpions, for instance, we have median 

 cement-glands [cf. the stigmatic combs of Galeodes) with lateral stigmata in the 2nd-3rd abdominal 

 segments, and in the Spidery we have median and lateral spinning-mamillse. In all, we can comit 

 five consecutive segments of the abdomen with spinning- or cement-glands near the median line (i. e. 

 if the spinning-glands of Phrynus are median and do not belong to the distal ends of the limbs), with 

 lateral stigmata and spinning-glands occurring at the same time and on the same segments. 



These four series, two on each side, can, it seems to me, only be deduced in different ways from the 

 setiparous areas on the parapodia of the original Annelidan ancestor, as shown in fig. 18 B (PI. XXXIV.). 

 This will explain the present variations of positions and the serial arrangement of the glands in Arachnids 

 in a very simple manner. 



To this point I shall return in the section on the Phylogeny of the Araehnida. 



