102 VARIATION IN LEPTINOTARSA. 



study of variation, whether individual or geographical, or for whatever pur- 

 pose, but especially in evolution studies, one may be easily led into serious 

 errors by the gathering of material here and there, and at one time only. 



We must first discover whether this phenomenon exists in the form chosen 

 for study, by observations covering several generations, and if it is found 

 its limits must be determined as near as possible. This variation may be 

 expressed by any of the variation constants, or by oscillations of the empirical 

 mode expressed in deviations in classes from the empirical race mode, i. e., 

 the empirical mode of the species. 



The causes of place variation in any species are the fluctuations from time 

 to time of the individual factors in its environmental complex. These are 

 moisture, temperature, and food commonly, and other factors less frequently. 

 It is possible in many cases to determine rather exactly the cause of a given 

 place variation, but the determination is not of any great value. Thus from 

 the data given it is possible to show that excessive variations were found 

 during this or that deviation from the normal. 



As far as I am able to determine from observations and experiment, place 

 variations result in no permanent modifications, nor do the changes seem to 

 be inherited. They are produced in these beetles during the larval and pupal 

 periods, are somatogenic, and purely fluctuating. Place variation results fre- 

 quently, as can be seen in the data given, in the production of prophetic and 

 historic skewness ; indeed, in the material which I have studied I believe that 

 most of the cases of skewness are due to this cause and do not have the 

 meaning that might be attributed to them. 



This place variation must necessarily be a troublesome factor in the study 

 by biometric methods of evolution, geographical variation, or selection. In 

 my own work it has been the rock upon which many cherished schemes have 

 been wrecked, and I suspect it has not yet completed its destructive work. If 

 one would study any of the broader problems of evolution by biometric meth- 

 ods he must first of all determine whether in the material chosen for study 

 this phenomenon exists, and if it is found, too great care or too long a time 

 can not be spent in the elimination of this factor. At present I know of but 

 one method of doing this that is, by collecting data and material over a suffi- 

 ciently long period to determine the range of this form of variability. Un- 

 happily this demands time, patience, and often funds which the investigator 

 will not or can not afford. The failure to take into consideration this place 

 variation vitiates the validity of a large part of the biometric work that has 

 been done, and there is no reason to think that it will be otherwise in the 

 future. The consideration of this phenomenon will in a very large percentage 

 of cases make biometric work and the statistical study of variation and evolu- 

 tion efforts of much longer duration than hitherto, but the demonstration of 

 the existence of place variation in no wise vitiates the results which may be 

 obtained by biometric methods when this phenomenon is given due consider- 



