Phylogeny of the Arachnida. 307 



sions. The endosternite, from which sprang the dorso-ventral 

 bundles of the cephalothoracic muscles, was strongly deve- 

 loped. The Protarachnon was a littoral form, which passed a 

 portion of its life in the water and the remainder upon the 

 land. Oviposition took place in the water; the development 

 was accompanied by a post-embryonic metamorphosis. 



From the above description it is evident that the hypo- 

 thetical Protarachnon must have differed considerably from 

 the primitive type of the other Tracheata. This diff^rence 

 must be acknowledged if the relationship of the Xiphosura to 

 the Arachnida be assumed, since, according to this latter 

 view, from the forms closely allied to the Protarachnon arose 

 the branch of the Gigantostraca, which gave origin to the 

 Xiphosura as a lateral offshoot. Consequently the ancestral 

 form of the Aracknoidea tons an Arthropod^ that of the group 

 Peripatus + ]\[yriopoda + Hexapoda^ however^ a Peripatus-like 

 Land-Annelid. This latter form differed but little and merely 

 in a quantitative respect from the ancestral form of the whole 

 of the Arthropoda. From the primitive Crustacea the existing 

 Crustaceans as well as the Trilobites have been developed ; 

 the Protarachnon and the Gigantostraca, however, are a side- 

 branch of the latter that split off at the commencement of or 

 even before the Palaeozoic era. The relation of the Protarach- 

 non and the Gigantostraca to the Trilobites distinguishes 

 the view expressed above from that of E,ay Lankester. 

 Without touching upon the relations of the Arachnida to the 

 Crustacea, I will merely remark that this view differs alto- 

 gether from that of those authors, who separate the Arachnids 

 indeed from the rest of the Tracheata, but do not derive the 

 Crustaceans from well-segmented Annelids (Fernald, No. 15, 

 pp. 493-499 ; Oudemans, No. 51 ; and others). To me the 

 paper by Oudemans (No. 51), in which much attention is 

 devoted to the Acafina, is interesting. The author seeks to 

 show the artificiality of the Arthropod type, and demonstrates 

 the necessity of dividing it into certain independent groups. 

 In his opinion the Trilobites + Xiphosura -I- Gigantostraca + 

 Arachnoidea (not including Acarina and Tardigrada) form a 

 perfectly separate group which commences with a hypothetical 

 form, Proagnostus. This Proagnostus was ihe larval form of 

 the whole of the Arachnida {loc. cit. p. 51) ; it differed from 

 the Nauplius chiefly in the fact that it consisted of at least 

 six fused postoral segments, and that the formation of the 

 following segments proceeded far from the posterior end of 

 the body. The conclusions of this naturalist are based only 

 upon observations upon the metamorphosis of the palseozoic 

 Trilobites {Agnostus, Phacops) ; yet at this epoch the Gigantos- 



