2 BULLETIN 668, U. S. DEPARTMENT OF AGRICULTURE. 
50 to 75 per cent of the plants. To thin out large numbers of sur- 
plus plants, each of which probably would be capable of producing 
practically as many bolls as any of the plants left in the row, is 
extremely wasteful when the potential value of every seed that 
could be produced is considered. 
The planting of so many cotton seeds in a hill has still another 
disadvantage from the standpoint of increasing valuable selections. 
The supply of seed when planted in the usual way often does not 
permit planting in more than one place, and the destruction of that 
planting, resulting from cold, hail, or any other cause, may mean 
the total loss of the progeny. A method making possible duplicate 
or triplicate plantings would reduce that danger of loss and at the 
same time afford an opportunity for studying the progeny under 
different conditions of soil and climate. 
The waste of seed continues in all the later plantings of the seed 
produced in the progeny row. If the quantity is sufficient for a 
regular field planting, a mechanical planter is likely to be used, the 
seeds being drilled in as many rows as the supply will permit. The 
seeds thus planted produce under favorable conditions many more 
plants than it would be advisable to leave to mature. In thinning 
to the desired stand, 10 or more plants may be destroyed for every 
one left, depending upon germination, season, and spacing. 
The loss thus experienced by the breeder during the three or more 
years that are required to obtain in quantity seed of his selection is 
enormous and makes plain the need of having less wasteful methods. 
Such methods would have a special importance in connection with 
the breeding of new varieties and would be of even greater practical 
importance in helping to maintain the uniformity of superior strains 
by continued selection. Methods that make possible a more rapid 
increase of select cotton would also be of special value to farmers, 
who do not. as a rule, practice selection because the process of produc- 
ing sufficient quantities of pure seed appears to require too much time. 
With the need of improved methods in mind, experiments were 
conducted in 1917 at the United States Experiment Farm at San 
Antonio. Tex., in cooperation with the Office of Western Irrigation 
Agriculture, the usual methods of increasing select stocks of cotton 
being compared with others in which a number of different nurse crops 
were used. The object of these nurse plantings was to determine 
whether as good a stand of select cotton can be secured where other 
kinds of seed supplied the lifting force necessary to break the soil 
crust as where only select seeds were planted. Could this be done, 
there obviously would be a material saving of select seed. 
To discuss the observations made during the progress of these ex- 
periments is the purpose of this bulletin. 
