COTTON-GROWING 165 



the departure of the flowering curve from a theoretical 

 form gives evidence as to the nature and magnitude of 

 the causes affecting the plants. 



The utility of these Plant-Development Curves is 

 almost endless, as the author has shown in Egypt. To 



study the data obtained in ordinary field 

 > d P ex P er i ments > a ^ ter being accustomed to using 



these continuous records, is like making use 

 of a dictionary from which many pages have been torn 

 away. The difference between them and the ordinary 

 data for three pickings is similar to the difference between 

 the inked trace of a barograph and the guesses founded 

 on the tapping of the barometer. The trouble of ob- 

 taining them is appreciable, but the cost is trifling; one 

 experiment with 100 plots on 2 acres, conducted by 

 the author, in which many more data were taken than 

 those sketched above, cost 30 more than the ordinary 

 cost, for all observation salaries, stakes, headman's time, 

 and clerical appliances. Three sets of plots of fifty each, 

 directed to the examination of ten different spacings, 

 sowing-times, and manurial arrangements, should, a year 

 later, repay the outlay upon them many times over. 



A modification of these methods may be used, with 

 certain limitations, for testing varieties or strains when 



only very small amounts of seed are avail- 

 ^ ai able. A field of ordinary crop is taken, and 



in it are sown rows of the seed to be tested, 

 replacing the ordinary crop seed which should have 

 occupied the same places. These rows consist of about 

 a hundred holes each, and not less than five such rows, 



