scales. Aspidiotus s. str. is of course the type of the entire 

 genus, and for the present purposes may be regarded as the 

 most generahzed. Diaspidiotus represents the extreme of 

 speciahzation in one direction. Hcniiberlesia is the inter- 

 mediate group. Of the former, A. osbonii and Jmnteri are 

 of the extreme type. The resemblance to the general form 

 is more and more marked in the various species in the order 

 indicated in the key, wi\\W. jatropJiac ?iwd punicac are reached, 

 when punicae presents characters such that it cannot with 

 certainty be placed in this sub-genus or in Hemiberlesia. It 

 is an intermediate form and Diaspidiotus grades impercept- 

 ibly into Hemiberlesia. Not distantly related to punicae is 

 convexus, placed in Hemiberlesia, all the species of which 

 show closer relations to Aspidiotus s. str. in their increased 

 fringe of plates and the absence of chitinous processes, or their 

 replacement by broad though thin chitinous bands. The 

 greater number of these species are lower austral or tropical 

 and in their distribution we see again their approach to the 

 regular forms of the genus, i. e. Aspidiotus s. str. 



Throughout, the characters of the female anal segment 

 are regarded as the principal ones for classification as it has 

 been repeatedly shown that scale characters in closely related 

 forms are unreliable as diagnostic characters. Though val- 

 uable as specific characters in many cases, they cannot be 

 considered as the result o-f relationships, but rather of en- 

 vironment. 



To Professors Osborn, Bruner, Webster, Mally, Rolfs, 

 Gossard, Parrott, Hunter, and others who have so kindly 

 furnished specimens and other assistance, the author wishes 

 to express his hearty appreciation. To Prof. Summers, of 

 the Iowa Station, thanks are due for many valuable sugges- 

 tions and assistance in the way of literature. To Prof. T. D. 

 A. Cockerell the writer is most deeply indebted. Briefly 



