THE MOSAIC DISEASE OF CUCURBITS. 53 
tion in out-of-the-way places. All such infection is so easily possible 
and so difficult to trace that field observations are always subject 
to doubt as proof of seed transmission of the disease. 
TESTS WITH SEED FROM MOSAIC PLANTS. 
In order to obtain more definite data in regard to seed transmission, 
a number of tests have been made with seed saved from mosaic 
plants. These trials were conducted both in the greenhouse and in 
the field, the field tests being at Big Rapids, Mich. 
Tests of 1915. — Seed collected from mosaic plants at Hamilton, 
Mich., during the fall of 1914 was tested in the greenhouse of the 
department of botany at the Michigan Agricultural College at East 
Lansing, Mich., during the winter of 1914-15. The seed from indi- 
vidual plants was not kept separate in these trials, most of it being 
very immature, and only 90 plants were obtained from the entire lot- 
These were grown in pots on sterilized soil for five weeks, but no signs 
of mosaic developed. 
Tests of 1916. — During the fall of 1915, seed was collected from 
mature fruits of mosaic cucumber plants, using only those fruits 
which actually showed evidence of the mottling and deformity charac- 
teristic of the disease. The seed from each fruit was removed sepa- 
rately, and in planting these separate lots were kept distinct. 
A preliminary test was made during the winter of 1915-16 in the 
greenhouse at East Lansing, Mich., 500 plants being grown in pots 
on sterilized soil at a temperature of 28° C. (82° F.) and kept under 
close observation for seven weeks. No evidence of mosaic appeared 
on any of the plants during this trial. 
An extensive field experiment was made during the summer of 
1916 at Big Rapids, Mich., on land which had not grown cucumbers 
previously and in a district nearly free from the disease. The seed 
was from the lot used in the greenhouse test during the winter and 
represented seed collected from mosaic plants in Michigan, Indiana, 
Wisconsin, Iowa, and Ontario, Canada. The seeds from each fruit 
were planted in separate rows 3 feet apart, with the seeds about 4 to 
8 inches apart in the row. Approximately 100 fruits were represented, 
which gave about 4,500 plants for observation. The germination 
was very uneven, but this seemed to result from the immaturity of 
some of the seed rather than from any effect of the disease, since all 
seed which germinated produced plants of healthy appearance. 
Observations were made practically every day in the early part of 
the season and at least once every two days in August and September. 
All plants remained healthy and normal in appearance and no disease 
appeared in the adjoining plats until July 28. On this date mosaic 
appeared on uncaged plants in a plat close to the seed test, through 
the accidental escape of aphids used in other experiments. In spite 
