274 Journal New York Entomological Society, [^oi. xxiv. 



antennae are usually well developed and are composed of many seg- 

 ments in most insects belonging to this section. When the caudal 

 filaments are short, the antennae are usually correspondingly reduced. 

 Protrusile sacs occur on certain of the abdominal sterna in some 

 cases. Tarsi two or three-segmented, and abdomen with traces of 

 eleven segments. Maxillary palpi four or five-segmented. Ab- 

 dominal ganglia eight in number. Some of the families belonging to 

 this section are the Machilidse, Lepismatidae, Gastrotheidce, etc. 



The relationships of these three lines of descent to the higher 

 insects, and to the other Arthropods are rather puzzling. It may be 

 remarked, however, that the Proturon type exhibits indications of 

 affinities with such Plecoptera as Capnia, Leuctra, etc., on the one 

 hand, and with such Crustacea as Bathynella and other Anomostraca 

 on the other. The Rhabduron type offers suggestions of affinities 

 with the Plecoptera (and the closely related Dermaptera) and with 

 the Symphyla such as Scolopendrella. The Thysanuran type ex- 

 hibits indications of a relationship with such Plecoptera as nymphs 

 of Peltroperla on the one hand, and with certain Crustacea, such as 

 Ligia, on the other. Since the Crustacea, " Myriopoda " and Insecta 

 are so closely related, hov^^ever, it is merely to be expected some of 

 the Apterygota would retain characters suggestive of a relationship 

 to the Crustacea while others show a marked relationship to the 

 " Myriopoda," since all are connected by mutual bonds of relation- 

 ship, and members of the Apterygota would naturally retain char- 

 acters suggestive of both Crustacea and " Myriopods." 



The Rhabduran section is, on the whole, more closely related to 

 the Thysanuran section than to the Proturan section, but all three 

 lines are rather sharply demarked and apparently represent three 

 distinct lines of development, just as there are similar distinct lines 

 of development in the Pterygotan insects. These three sections, too, 

 exhibit differences of a greater value than that of distinct orders. 

 Indeed, there is a far greater difference between Machilis and Cam- 

 podea, than there is between Perla and Phasnia, and the difference 

 between Machilis and Lepisma (or Nicoletia) is as great, or greater, 

 than that between Mcropc and Sialis. Furthermore, if the latter 

 examples represent distinct orders, then Machilis and Lepisma should 

 be placed in distinct orders, to be consistent. The value which one 

 will attribute to these differences depends upon the individual him- 



