STATES RELATIONS SERVICE. 591 



in which copra meal was substituted for part or all of the grain feed, 

 appear to indicate that copra meal can not be substituted for the 

 entire grain ration, but it can replace a portion of it, probably up to 

 50 per cent. With cattle, copra meal, as a portion of the concen- 

 trate, has given economical results, since it can be secured at a very 

 low price. The use of a small amount of concentrate with cows on 

 an exclusive green-forage ration was followed with an increase in 

 milk throughout the entire period of lactation. With swine, a ra- 

 tion consisting of two parts cooked breadfruit, one part damaged 

 rice, and one part copra meal, together with 3 to 4 ounces tankage 

 per head daily, proved a satisfactory ration for dry sows and mature 

 boars during a feeding period of 125 days. 



The poultry work has been reduced to experiments in breeding, 

 particularly an attempt to develop a new breed through the crossing 

 of the Rhode Island Red with a selected strain of white native 

 hens. The first cross has resulted in more vigorous and livelier 

 fowls than the Rhode Island Reds but of poor egg-laying qualities. 

 All sorts of colors of plumage were obtained. In the second gen- 

 eration crosses some pure white birds have been secured. The work 

 with purebred Rhode Island Reds is progressing satisfactorily and 

 through continued selection a flock of high producing fowls has 

 been developed. Surplus breeding stock was disposed of through 

 boys' and girls' poultry clubs. 



In the forage-crop investigations, grazing and feeding tests were 

 made with Napier, Guatemala, and Para grasses, and Japanese cane. 

 It was found that Napier grass made the quickest growth, and that 

 it should be cut when from 36 to 45 inches in height, otherwise the 

 stems become woody. Guatemala grass remains succulent much 

 longer and may be cut at a much later stage. The season was favor- 

 able for the growth of grasses and large yields of green forage were 

 obtained, 83 tons per acre for the third cutting of Guatemala grass 

 and 68,5 tons for the second cutting of Napier grass. Japanese 

 cane gave 87.5 tons for its second cutting. 



The corn work was confined to improvement. A four-year selection 

 of Guam white corn yielded at the rate of 26 bushels per acre. A 

 small 5^ellow variety received from the Hawaii Experiment Station 

 yielded at the rate of 41 bushels per acre. Time of planting, 

 variety, and fertilizer tests with rice were continued with varying 

 results. Early planting and the application of nitrogen in the form 

 of ammonium sulphate have given the highest yields. A number 

 of ranchers near the station were induced to try early planting of 

 rice, and yields of as much as 64 bushels of unhulled rice per acre 

 were obtained. 



The soil investigations were continued, the effect of continuous 

 cropping and rotation of crops being studied. Pot experiments 

 with old and new soils showed that grasslands of Guam newty 

 brought under cultivation are decidedly acid, and there was an im- 

 mediate response to tlie application of lime. 



Work with tropical fruits consisted merely of keeping the plant- 

 ings in as good condition as possible and propagating the better 

 varieties for further planting. 



The vegetable work has been reduced to a very small area, the 

 most of the garden tract having been planted to cowpeas. 



