358 THE VIEWPOINT OF A PHYSIOLOGIST 



general, nutritional requirements are so similar and so general 

 that subspecific differences are slight. Selection of specific food 

 plants by herbivorous insects appears to be based on taste prefer- 

 ences for certain specific organic components of the plants which 

 are repellent or neutral to other insects. Thorpe (1939) pointed 

 out that selection of food plants can be developed by habituation, 

 as by repeatedly exposing the larvae to a particular food, and he 

 has experimentally habituated Drosophila to a medium contain- 

 ing peppermint and the ichneumid Nemeritis from Ephestia to 

 Meliphora. Dixippus morosus was "forced" to feed on ivy, initially 

 an unfamiliar food, instead of privet; initial mortality was high, 

 but after several generations the progeny accepted the ivy readily 

 (Sladden, 1934). Eventually natural selection can result in ge- 

 netic fixation of food preferences which are initially phenotypic 

 variants established by habituation. A population of sawfly, Pon- 

 tania salicis, which had been feeding on Salix andersonia, was 

 forced to eat only S. rubra, until after some four years the saw- 

 flies failed to make galls on S. andersonia when this species was 

 planted near-by (Harrison, 1927). This interesting experiment 

 deserves confirmation under carefully controlled conditions. 



Most spectacular are tolerances of natural alkaloids by some 

 insects, such as those feeding on tobacco, and the tolerance of 

 modern insecticides by various strains. In some insects, resistance 

 to DDT seems to be carried by a single gene but in most it is 

 multifactorial. In many the genetic pattern necessary for resist- 

 ance was present in the population before the specific stress of 

 insecticides appeared (DDT resistance in Musca, Lichtwardt et 

 ah, 1955 ) . In a few the resistance appeared during many genera- 

 tions of selection ( Crow, 1954 ) . The complexities of resistance to 

 insecticides have been recently reviewed (Metcalf, 1955). Re- 

 sistant strains do provide a basis for natural selection of a particu- 

 lar race, which may ultimately lead to speciation, since areas 

 where the toxin exists are spatially separable from those without 

 the toxin and are well maintained. 



Examples of species with very minor morphological differences 

 but isolated by food preferences are: the homopterans Psylla 

 mail on apple and peregrina on hawthorn (Lai, 1934); and two 



