50 JOURNAL OF ECONOMIC ENTOMOLOGY [Vol. 9 



desert their proper secondary host, evidently instinctively trying to 

 find vegetation at a distance from their sister migrants where their 

 progeny will not lack for ample supplies. Others, which may be too 

 restless to work with in bright daylight, become docile at dusk. 



For these and many other reasons it becomes evident that a failure 

 with a migration test gives no data. 



If an investigator fails in one hundred attempts to colonize thistle with 

 migrants from plum that will not be a safe reason for him to conclude 

 that he is not working with Aphis cardui, or that this thistle aphid has 

 nothing to do with the leaf deformations of the plum in the spring. 

 It has been my own experience that negative data with aphids under 

 such conditions are just no data at all. If the structural characters 

 are such as warrant the migration test in the first place, they warrant 

 a patient continuation even in the face of repeated failures. 



On the other hand (and this is the most encouraging and stimulating 

 circumstance in connection with aphid migration tests), a single suc- 

 cess goes a long way to prove the case. Barring complications, a 

 single success is enough, and repetitions and verifications are needed 

 only as safeguards in that respect. For these insects are remarkably 

 stable as to their exclusive tastes in vegetable juices and a given species 

 will die before it will submit to the sap of any plant not on its approved 

 dietary. So if the progeny of the migrants accept the food plants 

 given them in the laboratory to the extent of developing upon it 

 from the first instar to maturity, it is safe to conclude that that food 

 plant is one which they would accept in the field under favorable con- 

 ditions, even though, with the wider choice of the open, a different 

 one might be given preference in certain localities. Such proof 

 should rest with the behavior of the progeny of the migrants and not 

 with the migrants themselves, for the migrants, as has been sug- 

 gested, have man}^ ways of tantalizing the hopeful investigator. 



Since the real proof of the validity of a tested food plant rests with 

 the ability of the progeny of the migrants to develop upon it, it is much 

 simpler to work with the spring migrants than the fall, return forms 

 when dealing with the Pemphigini for the reason that it is easier 

 to be sure that the immediate progeny of Pe7nphigus bur sarins, for 

 example, are developing upon the roots of lettuce than it would be to 

 be -sure that the stem mothers causing bur sarins galls in the spring 

 are hatched from eggs deposited by the progeny of the return migrants 

 from lettuce the fall before. Aside from the fact that there are 

 likely to be fewer complications with, the spring forms, with many 

 species it is often easier to locate and obtain abundant material in 

 the spring. 



