350 



JProceeding-i of the Royal Iriih Academy. 



grouping them "vrith. the Axtkropoda, since they possess the more 

 fluidamental characters of the Phylum — the ostiate heart, the peri- 

 cardium and secondarily formed body-cavity consisting of swollen 

 blood-spaces, and the greatly reduced coelom (Sedgnict, '87). In 

 their soft skin and simple segmentation, these animals strongly recall 

 the Annelids. Only three pairs of limbs ^or two, if the feelers be, as 

 believed by some, primitively pre-oral) are carried on the head, instead 

 of the five usual in Arthropoda. But perhaps the most striking 

 feature in which the ilalacopoda differ from other Arthropods, is the 

 presence of paired ccelomic excretory ducts ia all the body-segments. 

 These have been constantly compared with the nephridia of segmented 

 worms. 



The number of body-segments varies greatly ia the different genera 

 of the !3Ialacopoda ; and it is hard to determine whether the ancestral 

 stock of the class had few or many segments. Bouvier ('00) regards 

 the Tropical American genus Peripatus, with from twenty-two to 

 over forty pairs of legs, as the most primitive, poiating out that in this 

 genus the genital opening is situated between the penultimate legs, 

 while iu most of the other genera it is between the hindmost existing 

 pair, those corresponding with the hindmost ia Peripatus having been 

 lost. But ia their method of development, the Australian Peripatoidinae, 

 with their large, yolked eggs, in some cases hatched outside the body 

 of the mother, are certainly the most primitive members of the class, 

 the views of Sedgwick ('88), adopted by Korschelt ('99), beiag much 

 more reasonable than the suggestion of ~Willey ('98) that the acq^uisi- 

 tion of yolk in this class has been recent. It is very iateresting 

 to note, then, that the segmentation of these anioials agrees closely 

 with that of the typical Arthropods. They have from fourteen to 

 sixteen pairs of legs, so that, allowing for the head appendages 

 and the lost pair of legs on the post-genital segment, we arrive at 

 from eighteen to twenty limb-bearing segments. As regards their 

 segmentation, therefore, the Malacopoda might have been derived 

 from the typical Arthropodan stock, although the number of segments 

 in the class is too variable to justify any definite theory on this 

 subject. And such very primitive characters as the set of paired 

 segmental organs and the simple nature of the eyes, obliges us to 

 consider the Malacopoda as an ofEshoot from the far-off ancestral stock 

 of the other Arthropodan classes. As poiated out by Sedgwick ('95) 

 and Lankester ('97), they stand far below the rest of the Arthropoda. . 

 Any attempt to derive the Insecta directly from them, through the 

 Chilopoda, is vain ia view of the niunerous correspondences between 



