38 The Ottawa Naturalist. [Jtme-July 



The Cynipidae or gall wasps present an entirely different 

 condition so far as the relation to the flora is concerned. They 

 attack plants referable to only six botanical families, and assign- 

 able to but eleven plant genera. There is, however, a most 

 striking concentration in food habits, since a very large pro- 

 portion of the more than 300 gall makers subsist at the expense 

 of the Fagacese which, for this group, means the genus Quercus, 

 the exact number in our list being 277, though this figure, like 

 "those above, is an approximation. Thirty species have been 

 rreared from the Rosaceae, 21 (Rhodiles) living at the expense 

 of the genus Rosa. The other species referable to the Cynipidae 

 are scattered in their food habits, the most evident concentra- 

 tion, and this far from marked, being the 12 species reared from 

 various Compositae, the genera Silphium and Lactuca producing 

 four and three, respectively. This marked limitation in food 

 habits is accompanied, as might be expected, by a high degree 

 of specialization in structure. 



The Aphididae or plant lice live on a great variety of plants, 

 though the gall-making forms occur upon relatively few plant 

 families and genera, the most evident concentration in food 

 habits being in the genus Phylloxera, with its 29 species pro- 

 ducing galls on Carya. 



The nealy allied jumping plant lice or Psyllidae present a 

 similar condition in the genus Pachypsylla and its relation to 

 the numerous types of gall occurring upon Celtis. 



The occurrence of a number of galls produced by closely 

 related insects upon food plants belonging to a genus or even 

 species, indicates a physiological relationship, and some of these 

 groups at least offer excellent opportunities for the investigator 

 who would study the relation between the specific identity of 

 gall makers and the galls they inhabit. It is undoubtedly true 

 that marked diversity in gall structure usually indicates the 

 work of different insects, though there is a possibility that varia- 

 tions in the structure of these deformities may be related to 

 some extent at least, to the period when the infestation occurs; 

 in other words, oviposition before the tissues have swoollen to 

 any extent in the bud may result in a somewhat different de- 

 formation than if egg laying be delayed until the leaves are 

 partly unrolled. There are a number of cases where appar- 

 ently identical gall midges produce markedly different de- 

 formations in the same or closely allied plants, and we are in- 

 clined to believe that the time of infestation in relation to the 

 development of the host may be an important factor as well 

 as the part of the plant attacked. 



