OF THE ACARID^ TO THE ABACHNIDA. 281 



witli, however, this remarkable difference — some seven segments 

 posterior to the genital aperture and in front of the anal aper- 

 ture are missing. 



Now when we remember that in the development of all articu- 

 late animals the rule is that new segments are added in front of 

 the anal segment, we are justified in assuming, not only that the 

 anal segments are homologous, but that the first eight segments 

 in the Acarid are homologous with the first eight segments 

 in the Araneid. 



The segmentation of Tetranychus shows us very clearly, then, 

 that it ceased to develop new segments as soon as eight segments 

 had been formed, i. e. as soon as it had reached the typical posi- 

 tion for the genital aperture, whereas, in the Araneids, some 

 seven more segments are added between the genital aperture 

 and the terminal or anal segment. This is the first and most 

 important point, helping to bear out our suggestion that the 

 Mites are Arachnids arrested in their development. I may add 

 that the fact that no " waist " is formed in the Acarid between 

 the 6th and 7th segment, as in the Araneid, is a point of no 

 importance. A waist is simply a mechanical device to enable the 

 animal to bend its body, which the Mites as a rule do not require; 

 a few develop slight waists, but in another place, as will be men- 

 tioned below. 



It is of course not necessary to suppose that all the Mites are 

 to be deduced from the same Arachnida, or that all were arrested 

 exactly at the same stagre in their development. Some may have 

 developed more abdominal segments than Tetranychus. On the 

 other hand, it would be difficult for them to develop fewer. It 

 would be necessary to develop some portion of the abdomen over 

 and above the anal segment, as the genital glands typically arise 

 in this part of the body, and without the power of reproduction 

 the Mites of course could never have persisted as a distinct 

 family. 



Kramer has described the segmentation of a Mite {Alycus 

 roseus ?) with seven abdominal segments. Kramer counts nine 

 by reckoning the last two thoracic segments as abdominal. This 

 error arises from misunderstanding the arrangement of the limbs 

 so common among the Mites, two pairs pointing forwards and 

 two backwards, with a considerable space between the anterior 

 and posterior pairs, which is simply an adaptive modification. 

 The conclusion drawn from it by Kramer and others, that the 



