238 EXPERIMENT STATION RECORD. [Vol.40 



seed was first moistened sufficiently to cement the lint to the seed and then 

 placed in a 10 per cent salt solution, the big and medium seed sinking to the 

 bottom and the small inferior seed floating on the surface. The merhod is 

 said to be applicable to all kinds of cotton seed except that of the American 

 varieties, although certain modifications in the concentration of the salt solu- 

 tion are suggested for some local varieties. Additional data show the germina- 

 tion of the unseparated seed to have been 65.5 per cent, of the heavy seed 88.5 

 per cent, and of the light seed 41.5 per cent. Three-year average yields 

 amounted to 367 lbs. of seed cotton per acre from light seed as compared with 

 586 lbs. from heavy seed. 



Oil content of cotton seed as influenced by variety and selection, L. E. 

 Rast (Ga. State Col. Agr. Circ. 70 (WIS), pp. 4. fig- I).— Analyses of the seed of 

 48 varieties and strains of cotton grown in Georgia during 1916 are reported as 

 showing a range in the oil content of air-dry seed of from 17.64 per cent for 

 College No. 27 to 23.8 per cent for Mexican. For milling purposes the produc- 

 tion of seed with an oil content below 20.5 per cent is deemed unprofitable. 



Cotton production and distribution, season of 1916—17 (Bur. of the Census 

 [U. fif.], Bui. 1S5 (1918). pp. 1U. figs. I',).— Detailed tabulated statistics pre- 

 sented and discussed deal with the production of cotton from the 1916 crop; 

 also with the consumption. Imports, exports, and stocks of cotton; number of 

 cotton spindles; cotton seed received, crushed, and on hand; and with cotton- 

 seed products manufactured, shipped out, and on hand during the year ended 

 July 31, 1917. 



Cotton production and distribution, season of 1917-18 (Bur. of the Census 

 [U. S.] Bui. 1S7 (1918). pp. 1S5, fins. IS).— Statistical information similar to 

 that noted above is presented for the season of V.'17-18. 



South African fiber plants. — I, Ambari or Deccan hemp (Hibiscus canna- 

 binus). I. B. P. Evans (So. African Jour. Indus., 1 (1917), No. 8, pp. 198-208, 

 figs. 9). — H. eanvahinus. s:iid to be extensively grown In India and to be in- 

 digenous to Africa, is described as an erect annual which attains a height in 

 South Africa of from 5 to 11 ft, and which produces a coarse, harsh fiber 

 suitable for bugs, cheap cordage, yarns, etc. The possibilities of its production 

 in Africa are indicated. 



The cultivation of jute in Purnea, N. C. Chaudht-ki (Anr. Jour. Bihar and 

 Orissa [India], 5 (1917), A T o. 1. pp. 72-79). — The methods employed in growing 

 the crop and the preparation of the fiber for market in Purnea, said to be the 

 most important jute-growing district in the Provinces of Bihar and Orissa. are 

 described. 



Tests with lupines on sandy soil at Askov. 1894—1903, M. B.terre (Tidsxkr. 

 I'lnntravl. 25 (1918), No. 1. pp. lfS-55) . — The results of comparative tests of 

 yellow and blue lupines conducted for 10 years Indicated that the best yields 

 of seed were secured from seedings made the first half of April, and that the 

 blue lupine, as compared with the yellow lupine, yielded from 12 to 14 cwt. 

 more seed per tSndeland (1.36 acres) and blossomed and ripened from five to 

 six days earlier. 



The identification of varieties of oats in New York, E. G Montgomery 

 (Jour. Amer. Soe. A (iron., 10 (1918), No. If. pp. 17 1-17 If). — This paper, a con- 

 tribution from the department of farm crops, Cornell University, briefly re- 

 views work done by G. Stewart during the summer of 1917 in a practical test 

 under field conditions of the key to oat varieties prepared by Etheridge (E. S. 

 R., 36, p. 833). Samples of oats were obtained from New York State seedsmen 

 and from two experiment stations and were grown for Identification. 



Of the 22 varieties listed by seedsmen. 11 proved tO be of Swedish Select type 

 and 2 others to contain an admixture of Swedish Select ( >f 10 varieties ob- 

 tained from the experiment stations only 5 were definitely Identified as listed. 



