18 NORTH AMERICAN DIPTERA 



material and that every collection named means less time for institu- 

 tional work and, to make a blunt statement, it is only fair to expect the 

 institution to look for some return from the labor of its employees. 

 However, I do not know of anyone who is not happy to verify a 

 student's findings and to give opinions and suggestions concerning the 

 distinctness of specimens, and, of course, in such cases they are re- 

 turned if desired. 



I cannot condemn too strongly the professor who permits a student 

 to work on a systematic or biological paper and to allow or even urge 

 that the results be published unless the material has been examined by 

 a specialist, or the species of which the biology is being studied has 

 been identified by one in authority. Each year many students are given 

 such problems and the results are frequently very unsatisfactory. Too 

 much care can not be exercised in the preparation of scientific descrip- 

 tions of either genera or species and it should be realized that a good 

 collection must be available before any work can be faithfully under- 

 taken. 



The preparation of keys is a very imi)ortant matter and their in- 

 corporation in a contribution adds greatly to its value. Some authors 

 attempt, in their keys, to portray what they term a ''natural classifica- 

 tion ' ', but it should be borne in mind that this is impossible. We know 

 too little about the insects with which we are dealing to produce any 

 such result. In preparing keys we arbitrarily accept one or more 

 characters for the separation of groups and, if the key is to be useful, 

 the characters used for each se])aration must be clean-cut. In any 

 large group it will be found to be very difficult to find clean-cut char- 

 acters in every case, some forms apparently going into either group 

 although the character used may serve for the vast majority. In cases 

 of this kind the doubtful forms should be run through both categories. 

 In the so-called "natural" grouping it is very frequently necessary to 

 employ characters which are not conspicuous and such keys are liable 

 to prove very difficult to use. Characters expressing a degree of varia- 

 tion sliould be avoided wherever possible, i. e. front produced vs. frcMit 

 less produced. These characters may be obvious to the maker of a key 

 but are certainly not recognizable to other students whose collections 

 may be limited. It is usually easy to use a key if you know most of the 

 forms but often difficult or impossible otherwise. 



All keys should be dichotomous and the use of more than two 

 alternatives should always be avoided. In a short key one may use the 

 "a" and "aa" system but it is not to be recommended in any case. 

 Keys so constructed are more time consuming than such as are used 

 in the present work. In the present case an attempt has been made to 



