FIELD CROPS. 643 



had not swelled after 10 days were treated to overcouie the resistance of their 

 seed coats and nearly all of them after tliis treatment were capable of ger- 

 minating. 



Pentosans in the soy bean, G. Borgmesani {Stas. Sper, Agr. Ital., 40 (1907). 

 No. 2, PI). J] 8-120). — Analyses of 5 varieties of soy beans showed that the 

 Yellow Giant stood lowest in pentosans witli a content of 2.86 per cent, and 

 tht! blaclv soy bean highest with 3.80 per cent. The celhilose content of these 2 

 varieties was 4.6S per cent and 4.40 per cent, respectively. 



Sugar-cane experiments in the Leeward Islands, 1905-6, F. Watts et ai.. 

 (Imp. Dcpt. Agr. llr.v^ hidics, SiKjar-Caiie E.rpt.s. Lccirdnl Isl.. 1905-6, pt. 2, 

 pp. 54, dgms. 5). — This publication contains a report on fertilizer experiments 

 with sugar cane conducted at Antigua and St. Kitts in the season 1905-G. 

 The trials with plant canes have been carried on for G years and with rattoon 

 canes for 5 years. The results given are based on a repetition of each experi- 

 ment 58 times with plant canes and 39 times with rattoon canes. 



It is shown that commercial fertilizers are unprofitable for plant canes on 

 thoroughly tilled land which has received about 20 tons of good barnyard 

 manure per acre. The results appear so definite that it has been decided to 

 close this series of experiments with plant canes. 



The tests with rattoon canes have been divided into 2 series, the one, known 

 as the old series, which has already received an application of commercial fer- 

 tilizers as plant canes, and the other, the new series, which has not received 

 any commercial fertilizers as plant canes. The results on the old series of 

 tests, which has now been carried on for 5 consecutive years, show that nitrog- 

 enous manures are essential and ])rofitable in the successful growing of i"at- 

 toon canes. Nitrogen in a quick acting form, such as nitrate of soda or 

 sulphate of ammonia, should be given alone in one application at a fairly early 

 stage of growth. 



The new sei'ies of tests has been in jirogress but 1 year and no reliable re- 

 sults have as yet been secured. 



Cultivation of tobacco in Hawaii, J. G. Smith and C. R. Blacow (Hawaii 

 Hta. Bui. 15, pp. 29, pl.s. 3, figs. '/). — The discussions on the culture of tobacco 

 in Hawaii presented in this bulletin are based on the result of 3 years' work 

 undertaken to demonstrate the possibility of the production of tobacco on a 

 conunercial scale in the islands. Complete directions are given for the con- 

 struction and management of curing barns and seed beds and on the nn'thods 

 of held preparation, transplanting, topi)ing, suckering, harvesting, sorting, fer- 

 menting, baling, and marketing. 



The station reports the following yields per acre: Cuban tiller tobaccos 

 grown from seed brought directly over from Cuba, 600 to 1,200 lbs. of leaf, 

 Sumatra tobaccos from seed direct from Sumatra, 900 to 1,200 lbs., Cuban 

 tobacco from seed grown 2 or 3 years or more in Hawaii, 900 to 1,400 lbs., and 

 the second, third, and succeeding generations of Sumatra grown in Hawaii, 

 1,200 to 2,000 lbs. All of these types showed improvement in quality, and the 

 results of 3 years' worlc in Hamakua show a remarkaltle increase in the per- 

 centage of wrapper leaves the third year over that iiroduced by plants from 

 seed imi)orted direct from either Cuba or Sumatra. It is believed that with 

 full stands and a favorable season 30 per cent of either the Cuban or Sumatra 

 tobacco will produce high-grade wrappers, and that this percentage can Ite 

 materially increased. 



It is stated tliat the best tobacco districts in Hawaii are in the cloud belts 

 on the slopes of the higher mountains, wherever there is sufficient rainfall, and 

 that tobacco of good quality can be grown in the Hamakua, Hilo, Olaa, Puna, 



