July 1, 1920 Transmission of Mosaic Disease of Irish Potatoes 335 



August I. In spite of the differences in contact with diseased hills, the 

 progeny of the two lots were 27 and 35 per cent mosaic, respectively, 

 and the progeny of the three lots were from 25 to 35 per cent mosaic. 

 Aphid dispersal from neighboring mosaic plots was easy, and it appar- 

 ently nullified any effect that the difference in contact might have had. 



TEST OF SOIL HARBORING 

 GREENHOUSE EXPERIMENT 



At harvesting time in 191 8 one tuber was taken from each healthy 

 hill in two hill-selected lots. At Orono on January 14, 191 9, these tubers 

 were split with a flamed knife, and one set was planted in steam-sterilized 

 soil and the other in soil from which a mosaic plant had been removed on 

 December 30 or January 13. Nineteen pairs of half tubers were used, 

 and the plants from 7 pairs were mosaic by February 22, when from i to 

 20 inches high. The plants from the other 12 pairs reached their maxi- 

 mum height about March 5 and remained healthy until dug in April. 

 The second generation of the 12 healthy pairs was grown and found to be 

 entirely free from mosaic. It is clear that there was no transmission by 

 the soil in which mosaic plants had just been grown, all mosaic that was 

 shown evidently being transmitted by the tubers. 



FIELD EXPERIMENTS 



The greenhouse experiment described in the previous paragraph was 

 not concerned with certain factors in the possible soil-harboring of mosaic 

 in fields — namely, old stalks, volunteer potato plants, and insects. There 

 is no doubt, when the proofs of transmission by aphids are remembered, 

 but that volunteer mosaic plants may contribute to the infection of 

 healthy stocks planted where mosaic stocks were grown the preceding 

 season if they are not discovered and removed before the appearance of 

 aphids. Even if they are, other factors might cause the infection of 

 healthy plants. 



To test this supposition, three rows of Green Mountain stock from a 

 plot rogued in 191 8 were planted across the location of a 191 8 20 per cent 

 diseased Green Mountain plot and a wholly diseased one. Each mosaic 

 hill was dug and the seed piece examined. If volunteers are disregarded, 

 28 per cent of the 142 hills grown upon the ground of the all-diseased plot 

 were mosaic as were 28 per cent of the 481 plants grown upon the ground 

 which had produced the 20 per cent mosaic plots. This evidently was 

 from infection the previous season, since 27 per cent of the hills were 

 mosaic by July 15. 



A similar but more extensive test consisted in planting 19 rows of the 

 same stock across the ground which had produced 14 of the 191 8 plots. 

 Similar examination of the mosaic plants on July 30 showed 22 per cent 



