Carpenter — Relationships between Classes of Arthrojtoda. 847 



ducts open between the eigMli and ninth or on the ninth segnicnt ; 

 in the Symphyla between the ninth and tenth segments ; in the Mula- 

 costraca generally on the eleventh (female), and thirteenth (male^i ; in 

 most Insects between the sixteenth and seventeenth (female), and on 

 the eighteenth (male). (See Table, pp. 354-5.) 



From this survey a strong presumption arises that the median 

 position which, still characterises the Crustacea is primitive, and that 

 a very slight shifting forwards, or a more extended but gradual shift- 

 ing backwards, has led to the position of the apertm-es in the other 

 classes. According to this view, the common ancestors of Insects, 

 Centipedes, and Millipedes had the genital ducts opening about the 

 eleventh segment. In the Symphyla and Diplopoda their position 

 has been shifted forwards, in the Insecta and Chilopoda backwards. 

 The structure of the ovaries in the Thysanura suggests that the genital 

 ducts of Insects are not the representatives of some special pair of 

 segmental organs, but longitudinal mesodermal vessels, analogous to 

 the archinephric ducts of Vertebrates, into which the ccelomic seg- 

 mental ducts open. There is no unlikelihood, therefore, in the gradual 

 shifting far backwards of the reproductive openings among the 

 Insects. 



Relationship between Arachnids and Crustaceans. 



The discusion in a former section of this essay on the relationship 

 between the various orders of Arachnida led to the conclusion that the 

 primitive Arachnids were aquatic animals, breathing by means of 

 appendicular gills. ISTatui-ally, therefore, we compare the Arachnids 

 with the Crustacea rather than with the Insecta. The immediate 

 progenitors of the Arachnida appear to have possessed a head with 

 four pairs of limbs, a thorax with thi-ee segments, and an abdomen 

 with thirteen segments and a telson, only six of which can be clearly 

 shown by comparative morphology to have carried appendicular gills. 

 But embryological evidence enables us to postulate with confidence 

 still more remote ancestors in which the head carried well-developed 

 compound eyes and five pairs of appendages, while it may be supposed 

 that all the abdominal segments, except the anal, bore limbs. In these 

 very ancient Arthropods, all the limbs, except the feelers, hadambulatory 

 and branchial branches ; and one important f eatui-e in the evolution of 

 the Arachnida must have been the division of labour between the 

 anterior and posterior limbs, the former becoming specialised for loco- 

 motion, the latter for breathing. Another was the loss of the feelers 



