PLANT VIRUS PROBLEM— SMITH 347 



similarly with self-colored stocks [10]. The variegation in these 

 flowers has been shown to be due to a virus carried to the plants by a 

 species of greenfly from virus-infected broccoli or cauliflowers in the 

 neighborhood. 



In the writer's opinion viruses play a larger part in the production 

 of variegations in flower colors than is usually supposed. For in- 

 stance, inoculations from the petals of common variegated mauve and 

 white and mauve and yellow violas, picked at random from the 

 garden (pi. 2, fig. 3), to healthy tobacco plants of the White Burley 

 variety, produced in those plants a virulent mosaic disease. The 

 virus is also capable of infecting several other species of Solanaceous 

 plants. Experiment shows that the virus causing this variegation 

 is a strain of cucumber mosaic virus (cucumber virus I). 



Some of the mosaic viruses affecting ornamental plants may 

 produce little effect on the plant other than the change in the color 

 of the flowers. It is quite likely therefore that a systematic inquiry 

 into the question would show that other familiar flower variegations 

 may be due in part to virus infection. There seems, however, to be 

 a common element in the appearance of this type of variegation, i e., 

 a penciling or flecking of the colors and a break in the hard line 

 dividing two colors. 



The next question is the important one of how plant viruses are 

 transmitted in nature from diseased to healthy plants. The ma- 

 jority of plant viruses depend upon insects for their dissemination 

 from plant to plant, and this relationship between insect and virus 

 is one of considerable interest. The insects concerned in the spread 

 of plant viruses are nearly all of one type, a type of insect which 

 feeds in a particular way which seems to be well adapted for the 

 injection of the virus into the plant. These insects belong to the 

 order Hemiptera and are of the sap-sucking type. The method of 

 feeding of this type of insect is well demonstrated by means of 

 the photomicrograph shown in plate 2, figure 2. 



Insects are not merely mechanical vectors of the virus, but in 

 all probability some kind of obligate relationship exists between the 

 two. The following facts seem to bear this out : Certain viruses can- 

 not be transmitted from diseased to healthy plants except by the 

 agency of insects and often only by one species of insect or one type 

 of insect and not by other closely related species ; some insect vectors 

 having fed once upon a virus-diseased plant remain infective for the 

 rest of their lives without the necessity for further recourse to a 

 source of virus infection. This suggests that the virus actually mul- 

 tiplies in the body of the insect. Further, some insects do not become 

 infective until a minimum time has elapsed after feeding upon a 

 virus-infected plant. This is often referred to, perhaps on insufficient 

 grounds, as the "incubation period" of the virus in the insect. A 



