448 NATURAL SCIENCE. June. 



among spiders it possesses four pairs of spinning mammillae. While 

 the Mygalomorphae rarely have more than two pairs, the Arachno- 

 morphae always have three, and Pocock thinks that the extra central 

 spinning organ (cribellum), found in some families of the latter, 

 represents the fourth pair of spinners of Liphistius fused together. 



In this new classification, therefore, the archaic characters of 

 Liphistius are emphasised, and it is regarded as a survival of an 

 ancestral form of spider in which the abdominal segments have not 

 yet become fused, and still bear the spinners on their ventral aspect. 

 The animal is, to some extent, a connecting link between the spiders 

 and the scorpion-spiders (Pedipalpi). We are thus led to consider 

 the relationships which exist between the various orders of the 

 Arachnida — a subject also dealt with by Pocock in a later paper (2). 

 His principal contention is that if arachnids are descended from 

 ancestors which must have possessed a long series of nearly similar 

 segments, those modern representatives of the groups in which the 

 segmentation is best preserved must be the most primitive. On these 

 grounds he considers the scorpions as the lowest living branch of the 

 arachnid stem, in opposition to the views of Thorell and others, who 

 have regarded them as the highest. There can be no doubt that 

 Pocock's opinion will meet with general acceptance among biologists. 



Deriving the arachnid orders immediately from an ancestor with 

 a long abdomen of twelve segments and a telson, Pocock indicates 

 the modifications which have probably occurred in each group. In 

 the scorpions these segments are all present, but the hinder six are 

 much reduced in bulk to form the post-abdomen, the telson being 

 utilised as the sting. In the scorpion-spiders there has been a great 

 reduction of the hindermost segments ; the tail in Thelyphonns repre- 

 sents the telson. Liphistius, with its much-shortened and partially 

 segmented abdomen, connects these with the higher spiders in which 

 the abdominal segments are completely fused together. Nearly 

 related to the Pedipalpi, but with tracheae instead of lung-sacs, are 

 the SoHfugae ' and false-scorpions (Chelifers), with the number of 

 abdominal segments complete, or but slightly reduced. On a higher 

 branch of the stem bearing these come the harvestmen (Phalangida), 

 with the abdominal segments generally reduced in number, and fused 

 with the cephalothorax anteriorly. From these the mites (Acari), 

 with the whole body fused together, must probably be regarded as a 

 degraded offshoot. 



The relation of the mites to the rest of the Arachnida has been 

 also recently discussed by Bernard (3), who gives reasons for believing 

 that these creatures are not examples of true degradation, but that 

 they have become fixed at a larval stage of development. Comparing 

 their digestive, vascular, and nervous systems with those of the 



I Bernard {A. M. N. H. (6), vol. xi., 1893, pp. 28-30) has announced recently 

 the discovery of a sensory organ on the pedipalp of Pliiynus, similar to that on 

 the homologous appendage of GaUodes. 



