1893. THE CLASSIFICATION OF ARACHNIDS. 449 



higher Arachnids, he shows that they want the greater number of 

 the abdominal segments, while they possess, in a fair state of develop- 

 ment, organs homologous with those in the anterior part of the 

 body of the higher Arachnids. He speaks of the Mites as " larval 

 Araneids," but few naturalists will be inclined to regard them as a 

 direct offshoot from the spiders. Extreme smallness — one of the 

 supposed advantages gained by fixation at the larval stage — has been 

 acquired by many true spiders. Moreover, spiders, like most other 

 Arachnids, have no larval stage, but leave the egg in a form similar 

 to that of the adult animal. Whether degraded or arrested, the 

 affinities of the mites seem to be with the harvestmen, from which 

 group they are not readily defined. 



Trouessart (4), however, supports the view that, on account of 

 their development with a metamorphosis, the mites should form a 

 special sub-class of the Arachnida, and he separates the vermiform 

 mites, such as the Phytopti, from the rest as a separate order ; but 

 the larval stage of mites (in which they have but three pairs of limbs) 

 seems quite a secondary adaptation, for, according to Wagner (5) the 

 fourth pair appear in the embryo, to be afterwards concealed beneath 

 the skin of the larva and then to re-appear during metamorphosis. 



It is remarkable that the theory of fixation at the larval stage 

 has also been invoked by Von Kennel (6) to account for the origin of 

 the microscopic "water bears" (Tardigrada), which have generally 

 been regarded as degraded arachnids. He compares them with 

 certain midge-larvas [Cecidomyia) which have the power of partheno- 

 genetic reproduction, and suggests that if male organs were also 

 precociously developed a species might become permanently larval 

 in form. Though he does not assert that the Tardigrada are 

 " arrested " flies, the acceptance of his views — which certainly have 

 much to support them — would probably lead to the removal of the 

 group from the arachnids to the insects, in which the larval stage is 

 often of such supreme importance. 



The replacement of lung-sacs by tracheae in several groups of 

 arachnids opens up a question of much interest. Reference has been 

 made in Natural Science (vol i., p. 524) to the supposed origin of 

 the lung-sacs from the invagination or enclosure of gill-bearing appen- 

 dages. If this view be accepted, we must regard the abdominal 

 tracheae in the higher spiders, harvestmen, false-scorpions, etc., as 

 simplified lung-sacs. Pocock (2) suggests that increased lightness 

 would be an advantage derivable from this change ; but Bernard 

 (3, 7, 8) considers the tracheae the more primitive, and the lung-sacs 

 elaborated from them. He would derive the whole ventral series of 

 arachnid breathing-organs, spinning-glands, coxal glands, and poison- 

 glands from the ventral setiparous glands of an annelid ancestor. 

 The tracheae of insects, on the other hand, he believes to have arisen 

 from the dorsal series of setiparous glands, and he finds the hairy 

 area around the stigmata of the chrysalis of the Vapourer -Moth 



2G 



