222 KELLOGG AND BELL 



order the bee belongs) has been specialized by reduction, /. e. 

 where the veins have been reduced in number and degree of 

 branching and inter-connection by cross-veins, so that the 

 remaining vein framework is presumably in all of its details 

 essential to the best performance of the wing's function, namely 

 flight. This reliance on venation is not based on such theoreti- 

 cal grounds however, but on the practical one of experience and 

 wide observation. In many of these specialized insects the 

 venation is fairly uniform throughout a whole family, while 

 practically never are describable differences in venation ex- 

 pected to be found within generic limits. In Comstock's Man- 

 ual of Insects (the standard American systematic manual) the 

 keys to the families of Diptera and Lepidoptera are nearly 

 solely, and of Hymenoptera, largely based on venation. Thus 

 variation in venation is to be looked on as important. 



Measurements of Parts of Wing-veins. Wishing to be able 

 to give an accurate quantitative expression to some features of 

 the variation of the wing venation, we have measured certain 

 parts of veins whose limits are accurately established by sub- 

 tending cross-veins, or by the forking or branching of the veins 

 themselves. In order to determine the relation of any variation 

 in these measurements to the varying size of the whole wing 

 this size, as indicated by the measured length and breadth (or 

 when impossible to get the length owing to battered and broken 

 wing tips by the width, alone), has been ascertained for all of 



FIG. Ji. Fore wing of honeybee (drone) with letters indicating subtending 

 points on various veins. 



the wings studied. While at first thought the dimensions of 

 parts of the veins might be assumed to be directly related to, 

 t. e. 9 a simple function of, the size of the wings, yet an inspec- 

 tion of the parts selected for measurement will suggest the pos- 



