178 L. M. BLACK 



multipKes in its aphid vector. By using a refined injection technique they 

 obtained transmission of potato leaf roll virus from aphid to aphid through 

 15 serial passages while maintaining them on Chinese cabbage which is 

 immune to the virus. Their measurements indicated that the vims was 

 diluted 1/20 at each passage. Nonvirulrferous aphids failed to recover virus 

 from the Chinese cabbage plants on which the injected insects were main- 

 tained. These results demonstrated infectivity in the terminal passages, at 

 dilutions far in excess of those possible with the starting inoculum. The results 

 of Stegwee and Ponsen suggest that even the evidence for progressive decrease 

 of virus content in vectors following termination of acquisition feeding, for 

 increase of virus content in proportion to length of acquisition feeding period 

 and lack of interference between virus strains in the vector, may have to be 

 interpreted with even more caution than has been true in the past. 



Certain data that have been advanced as evidence against multiplication 

 are known to be unreliable. Black (1953b, p. 411) has pointed out that certain 

 criteria, such as loss of infective ability (Kunkel, 1926), irregularity of 

 transmission, or even failure to transmit during the whole life (Fukushi, 1940) 

 or for many weeks (Black, 1950) in the life of a virus-bearing insect cannot be 

 considered significant evidence for absence of multiplication, because these 

 phenomena are exhibited by some viruses that have been proven to multiply 

 in their vectors. Loss of virus followed by its reacquisition does not, of itself, 

 eliminate multiplication, because this may happen in both plant and animal 

 virus infections. 



Although the minimum noninfective period in the leafhopper following 

 acquisition of curly top virus is not so short that virus multiplication could 

 not have occurred, it is nevertheless in striking contrast to the minimum 

 incubation periods of about a week or more in all cases where multiplication 

 in the vector has been demonstrated. However, whether short and long 

 incubation periods are indicative of the absence and presence of multiplication, 

 respectively, as suggested by Black (1954), cannot be determined until the 

 accumulation of critical evidence on more cases. 



None of the critical tests applied to curly top virus has provided evidence 

 for multiplication in the vector. This indicates that certain properties that at 

 times have been attributed to multiplication can no longer be necessarily 

 so assessed. 



One of these properties is a high degree of vector specificity. In the case of 

 curly top, this specificity has recently played a dramatic role in revealing the 

 probable past history of the virus and vector. Only a single vector, Circulifer 

 tenellus, for the curly top virus has been found in North America. Oman (1948) 

 pointed out that C. tenellus had close relatives, not in the United States, but 

 in the Mediterranean area. Leaf hoppers apparently identical to C. tenellus 

 were found in North Africa, South Africa, Sicily, and Palestine (Frazier, 1953; 



