INSULAR EXPERIMENT STATIONS. 23 



it has been demonstrated that the proper time to apply fertilizers 

 for rice is before planting and not with the irrigation water during 

 the growth of the plant, as has been the practice. The rice plant 

 was found to take up its plant food early in its growth, and the stool- 

 ing or tillering was greatly increased by the early application of the 

 fertilizers. It has also been found that ammonium sulphate is a 

 better source of nitrogen for rice in Hawaii than sodium nitrate; in 

 fact, there appears to be some evidence that the rice plant, especially 

 when submerged, does not use the nitrates directl3\ During the year 

 the agronomist visited Japan and China for the purpose of studying 

 the rice industry in those countries. The Japanese, who make up 

 more than 40 per cent of the population of the islands, prefer Jap- 

 anese varieties of rice even though they cost more than the domestic 

 ones. To produce this type of rice, seed of about 150 varieties were 

 obtained for experimental purposes, and of four a sufficient quantity 

 of seed was secured to test their qualities when grown on a consider- 

 able scale in Hawaii. 



The experiments with cotton, begun three years ago, have yielded 

 such striking results that 500 acres of commercial plantings have 

 been made with every prospect of increased areas being devoted to 

 this crop in the coming year. It has been definitely determined that 

 cotton can be grown at any elevation up to 1,600 feet above sea level 

 and with a precipitation varying from 25 to 100 inches, although a 

 moderate rainfall gives the most favorable results. The experiments 

 are principalh^ Avith sea-island and Caravonica cottons, and as these 

 plants are perennials in Hawaii it is necessary to control their growth 

 to prevent their becoming too large. Experiments have shown that 

 the form of the plant and the time of crop production can largely 

 be controlled by pruning, and that the prunings can be utilized for 

 propagating. Although this method may not prove a commercial 

 success, it has proved valuable in the propagation of pure strains of 

 cotton where the possibility of crossing could not be eliminated. The 

 tendency to heavy production of cotton was shown in a field of sea 

 island, where the plants averaged TOO bolls at six months, and some 

 individuals bore as many as 1,200 bolls. This heavy load, especially 

 on the lateral branches, tended to bend the plants to the ground, in- 

 juring the lint, and as a consequence experiments are in progress to 

 secure plants with shorter limbs and of a more upright habit of 

 growth. The Caravonica plants are stronger and more upright in 

 habit of growth, and while the yield of lint during the first year was 

 very small, the second year a heavy crop was obtained. The strong 

 growth, greater ease of picking, and high percentage of lint make 

 this a desirable crop, and there is little choice between this and sea- 

 island cotton. Some strains of Egyptian cotton experimented with 

 grew vigorously and yielded better than sea island, and it is probable 



