No. 2368. AMERICAN SUBTERRANEAN GALLS ON OAK— WELD. 189 



attached, they are labeled as cotypes or paratypes in the collection, 

 but they are not listed as type material in the body of the paper or in 

 Museum type book. 



In conformity with this view, the authorship of a species must be 

 credited not to the one who first described and named the gall but 

 to the one who first described the maker of the gall. 



The number of specimens from which a new species has been 

 described may be seen at the end of each description where the 

 number of specimens measured to get size and range in length is 

 indicated. Type material of all the new species has been deposited 

 in the United States National Museum, and the number of speci- 

 mens so deposited may be seen under the heading " Type." The 

 balance has been retained by the author for reference or exchange, 

 for in many cases the material is sufficiently abundant so that 

 exchanges can be arranged with other museums or workers. When 

 the type material consists of specimens bred from different hosts 

 or from different localities or consists of both reared and captured 

 specimens, a "type" has been selected and the rest called "para- 

 types." When the series has been reared from a single polythala- 

 mous gall it is obvious that they are of equal value, and here the 

 term "cotype" has been applied. The term "cotype" is also used 

 for a series reared from a lot of monothalamous or polythalamous 

 galls all collected on the same host from one locality. Few errors 

 are likely to arise in this application of the term. 



The arrangement of genera here given follows that of Dalla Torre 

 and Kieffer in the 1910 monograph in Das Tierreich (Lief. 24), and 

 their usage has been followed also in numbering the segments of the 

 abdomen, calling the first free tergite of the apparent abdomen the 

 second, the first being fused with the first sternum to form the petiole. 

 The term parapsides is often here used for parapsidal grooves. Fig- 

 ures 23 and 25 are from negatives in the Division of Forest Insects, 

 eastern station : the rest are from photographs by the author. Unless 

 otherwise noted the galls are represented in natural size. 



The names used for the oaks are those of the seventh edition of 

 Gray's Manual for the northeastern United States, and for other 

 regions what seemed to be the best names available. Throughout 

 the paper the same name has been consistent!}^ applied to a given 

 oak, but the name used may not in all cases be the one on which all 

 botanists would agree. A study of the host relationships of the 

 gall-making Cynipidae will undoubtedly throw light on the relation- 

 ships of the oaks. Some species attack many oaks and others 

 discriminatingly confine themselves to a single kind. One rootgall- 

 forming species hero treated occurs on at least ten species, all in 

 the red oak group, and no doubt will be found on still others. One 

 Californian oak has over forty different galls upon it, none of which 



