334 Journal of Agricultural Research voi. xjx, No. 7 



were entirely healthy except for mosaic mottling in the few uppermost 

 leaves of several branches of a stalk in one hill. These leaves appeared 

 young. They had evidently been pushed hard against the inside of 

 the cage and had a very few aphid skins and aphids clinging to them. 

 They may have been infected as the result of contact before aphids 

 entered the cage, by aphids on the outside of the cloth against which 

 the leaves were pressed, or by aphids that came from mosaic plants in 

 the next row and that entered through a small hole that was found 

 to have been made accidentally in the cloth. 



FIELD EXPERIMENTS WITHOUT CAGES 



As was pointed out in a previous section regarding the test of the 

 seed-cutting knife, the mixing of all-mosaic stock and rogued stock in 

 two rows was not followed by a higher mosaic percentage for the rogued 

 stock than was shown by it in a control row. The negative results in 

 this case do not disprove the possibility of infection occurring too late 

 to be evident during the current season — that is, after the roots and 

 vines have become intertwined. 



In 1 91 8 five Green Mountain hill lots were found to be partly mosaic. 

 The healthy hills were harvested separately, were classified according 

 to their proximity in the row to a mosaic hill, and the tubers were 

 planted uncut in 191 9. Twenty-eight tubers were progeny of plants 

 each of which grew between two mosaic hills, and 54 per cent of them 

 were mosaic. Eighty-nine were progeny of hills each of which was be- 

 tween a mosaic hill and a healthy hill, and of these 63 per cent were 

 mosaic. On the other hand, 40 per cent of the 220 tubers from hills 

 each of which grew between two healthy hills were diseased. If these 

 2.20 tubers are arranged in five groups. Hi, H2, H3, H4, and H5, 

 according to the increasing number of healthy hills between the parent 

 and the nearest mosaic plant in the row, the groups contained, respec- 

 tively, 75, 53, 41, 33, and 18 tubers, with 56, 24, 54, 24, and 17 per cent 

 of them diseased. Since being next to a mosaic plant in the same row 

 seemed to increase the chance of infection as much as 54 or 63 per cent 

 is greater than 40 per cent, it evidently is a contributing factor in mosaic 

 transmission; but judging from the varying percentages of infection 

 among the classes of plants which were not next to mosaic hills in the 

 same row, it probably aids in the spread of the disease only by aiding 

 aphid transmission. 



A slightly different type of experiment consisted in comparing the 

 progeny of three small i-row Green Mountain lots, of from 100 to 200 

 hills each, from which the mosaic hills (respectively, 6, 16, and 30 per 

 cent) were removed on August i, 191 8, with two similar lots from 

 which the mosaic hills (respectively, 6 and 18 per cent), together with 

 each healthy hill next in the row to a mosaic one, were removed 



