216 THE CANADIAN ENTOMOLOGIST 



antennae are located higher up in the frontal region, between the 

 eyes. The eyes of Embiids and Grylloblattids are similar in out- 

 line, and do not extend upward on the sides of the head; while 

 in the typical Blattids, the eyes are more "reniform," and extend 

 for a considerable distance upward along the sides of the head. 

 The head is typically opisthognathous (i. e., mouthparts directed 

 backward) in Blattids, while in both Embiids and Grylloblattids 

 it is more prognathous (i. e., mouthparts directed forward). The 

 cervical scleritfes are astonishingly similar in Embiids and Gryllo- 

 blattids (as will be brought out in a paper now ready for publica- 

 tion) while these neck plates in both groups differ markedly from 

 those of the Blattids, and the same is true of the thoracic sclerites 

 in general. The bodies of the Embiids and Grylloblattids are 

 slender, while those of typical Blattids are broader and more 

 flattened. It would be possible to cite many more instances of 

 striking structural similarity between the Grylloblattids and 

 Embiids, (in points wherein both differ markedly from the Blattids) 

 but it is preferable to take up these structures point by point in 

 a series of papers in which they can be discussed more in detail, 

 and in which the points brought out can be illustrated by numerous 

 figures^ — which after all are more convincing than bare statements 

 of facts, since they enable one to form an opinion for one's self 

 concerning the features in cjuestion. 



The Grylloblattids are undoubtedly also very closely related 

 to the Phasmids such as Timema, and in all probability both 

 Phasmids and Grylloblattids were derived from forms quite similar 

 to the Embiids. The Phasmodidse (Phasmodes ravatriformis, 

 Westw'.) are insects very closely related to the Tettigonids ("Lo- 

 custidffi") which have retained many features suggestive of Phasmid 

 affinities, and beyond a doubt, a comparative study of Phasmodes, 

 Gryllohlatta and Timema, w^ould be extremely instructive from 

 the genealogical standpoint. It might also be mentioned that the 

 Phylliidae are too closely related to the Phasmids to be considered 

 as a distinct order (the "Phyllioptera") as I formerly believed, 

 since such Phasmids as Ectatosoma are quite closely related to the 

 Phylliids (as can be seen by comparing the males of Phyllium with 

 these Phasmids), so that I would now regard the Phylliids as a 

 sub-order of the Phasmid group, rather than as a distinct order. 



