100 REPORT OF THE AGRICULTURAL RESEARCH 



(4) Kil or Garo Hill cotton is more productive than 



the two last named, but our experience is that 

 it deteriorates rapidly to the standard of local 

 types. 



(5) Yellow flowered 'Neglecium. This yields a cotton 



of better staple. 



All the field experiments are on the white flowered or 

 Varhadi type of cotton. As in other parts of India, owing 

 to its hardiness, high produce and also high percentage of 

 cotton to seed, it is rapidly becoming the favourite variety 

 of the cultivators. Already Dr. Parr has 300 acres under 

 this in seed farms and next year he hopes to increase their 

 area to 3,000 acres. In price of crop it surpasses yellow 

 flowered cotton by Rs. 12 to Rs. 15 per acre. There is 

 little difference in the yield of broad cast and drilled 

 plants. The best spacing distance has been found to be 18 

 inches between rows and 9 inches in the rows. A yield of 

 10 maunds of seed cotton per acre has been obtained. 

 White flowered cotton grown with ground-nut yields about 

 half the ordinary unmixed crop. The general practice is 

 to grow cotton with Cajanus indicus. Dr. Parr says that 

 in the Aligarh District about half the cotton crop is irri- 

 gated. Southwards the proportion of irrigated cotton 

 decreases. 



Mr. Graham, the Manager of West's Patent Press at 

 Aligarh, thinks that, so far as cotton improvement is con- 

 cerned, it would be sufficient to increase the outturn. He 

 also thinks that a better staple should be arrived at. Few 

 traders I have met seem to appreciate the fact that an 

 increased staple in Bengals is invariably accompanied by 

 a falling off in yield and percentage of cotton to seed. 



After a discussion with Dr. Parr regarding his future 

 line of work we agreed that the following would be most 

 practicable : — • 



(1) To purify out the white flowered cottons and extend 

 their cultivation as pure types. They have 

 already been tested and proved to be more pro- 



