372 EXPERIMENT STATION RECORD. 



The grade of the slope of the Esopus Basin is 13.05 per cent or twice as steep, and 

 the fall of the river is 59.6 ft. per mile, or about 6 times as rapid as that of the Wall- 

 kill. The topography of the basin is also much more simple and direct. There are 

 no natural reservoirs in the Esopus Basin, while 42.52 sq. mi. or 5.04 per cent of the 

 total area of the Wallkill Basin is occupied by lakes, ponds, or swamps. 



The characteristic geological features of the Wallkill Basin are deep layers of fine 

 Loamy, marly soils of large absorptive capacity, while in the Esopus Basin there is 

 only a thin layer of soil which tends to shed water immediately rather than to 

 store it. 



Studying the weekly deviation of these 2 streams from 1901 to 1903, inclusive, it 

 was found that the percentage deviation from the average weekly mean flow was 

 78.24 per cent for the Wallkill and 83.69 per cent in the case of the Esopus. It is 

 thus shown that the combined influence of moderate topography, natural reservoirs, 

 and favorable geological conditions of the Wallkill are stronger in promoting even- 

 ness of stream flow than the compact forest < >f the Esopus. The difference, however, 

 is small, the forest cover overcoming in a very large degree the unfavorable effects 

 of steep topography, hard and dense surface rocks, and marked deficiency in natural 

 storage facilities. 



It is concluded, therefore, from the evidence in hand that temperature and pre- 

 cipitation being equal, the relative regularity of stream flow is determined first by 

 the topographic and geological conditions, and second by the character and extent of 

 forest cover. 



Fire protection in the teak forests of Burma, R. S. Troup {Indian Forester, 

 31 (1905), No. 3, pp. 138-146, dgms. 2).— The author made a careful study of the 

 growth and reproduction of teak on an area of 37 acres, which had had fire protec- 

 tion since 1872-73, and another area of 24 acres lying alongside, which had had no 

 fire protection and which was burnt over annually. 



From the data obtained it is shown that in moist, mixed, fire-protected forests the 

 bamboo grows so rapidly as to depress and severely injure teak trees, while with 

 annual burning the bamboo is suppressed sufficiently to permit of the very satisfac- 

 tory growth of teak. The locality chosen for this experiment is believed to be typi- 

 cal of hundreds of square miles in Burma, and it is believed that fire protection over 

 such areas is decidedly injurious to the development of teak. 



In Vol. 31, No. 7, pp. 383-388, of the same publication, a number of discussions on 

 the above paper appear, some of which confirm the author's views, and others point 

 out that in the very dry teak forests of Burma, where the cover is much lighter, fire 

 protection is of undoubted value in aiding the young teak plants to establish 

 themselves. 



Discovery of manicoba rubber forests in Brazil, H. W. Furniss {Mo. Consular 

 Rpts. [V. &], 1904, No. 287, pp. 52-57). — The author states that extensive forests of 

 "manicoba" or Ceani rubber {Manihot glaziovii) have recently been discovered in 

 the interior of Brazil, particularly in Bahia. 



Methods of manicoba rubber culture are discussed, with an account of milk pro- 

 duction and of extracting and collecting the latex. The quality of this rubber seems 

 to be very high. It is believed that it could be profitably introduced into the south- 

 ern part of the United States, and especially into Porto Rico and the Philippines. 

 Statistics are given on the exports of the Bahia rubber to different countries of 

 Europe and the United States during each of the years 1901 to 1904. 



Rubber tapping- in the botanic gardens, Singapore {Agr. Bui. Straitsand Fed. 

 Malay States, 4 (1905), No. 5, pp. 144-1*4)-— Tabulated data are given which show 

 the yields obtained from trees of different sizes tapped in the morning as compared 

 with those tapped in the evening. There were 15 trees in each experiment and 5 

 experiments were made. The trees in the different experiments varied in circum- 



