January 31, 1901] 



NATURE 



335 



the reflex observation is made first, and the instrument is then 

 turned in mid-transit to the direct position, in which the 

 exposure is of very short duration before the observation in 

 zenith distance is made. In the curve c, for which in each case 

 the direct observation was made first, the exposure in the direct 

 position would commonly be considerably greater ; and the 

 difference between this and the flexure-curve, attributable to 

 such exposure, is accordingly much greater than the same differ- 

 ences for the A and B curves. Note that the difference becomes 

 smaller under 40° zenith distance ; it would be zero in the 

 zenith in all cases. 



Other instances of apparent refraction within the tube are 

 found in the Greenwich observations : — 



When the north and south collimators are alligned by looking 

 through the holes in the telescope- cube, the collimation-error 

 obtained differs systematically from that obtained by alligning 

 the collimators with the telescope raised out of the way. This 

 can only be explained thus : one side of the instrument is com- 

 monly warmer than the other at the hour (8 a.m.) when these 

 observations are made, and the still air in the tube is affected in 

 like manner. If we suppose the air in the spaces A and B, 

 Fig. 3, to differ from each other 0° 5 Fahr. in temperature, and 

 to be separated by a surface which the path of the light between 

 the north and south collimators cuts at an angle of incidence of 

 80°, the light will be deflected o"*6o of arc, and the collimation- 

 error obtained would be in error by half this amount, viz., o"'30. 



/4 



■^o's p'akr: 



Fig. 3. — r. Shows path of light from north to south collimator refracted 

 within cube of telescope when air is unequally heated. 



which is about the difference between the values obtained by the 

 two methods. 



A similar discordance exists between the results for flexure of 

 the tube formerly obtained by raising the telescope and those re- 

 cently obtained by alligning the collimators through the cube ; 

 and tjiis discordance has a similar explanation. The flexure 

 obtained formerly, which I take to be the more correct, is 

 about o"-8o of arc, as employed above to explain the R - D 

 discordance. 



I conclude that means should be provided — and used— for 

 circulating the air in the tube when any observation— whether 

 of star, of collimation, or of flexure, or otherwise— is made with 

 a transit circle. I would also point out that the source of error 

 here considered is of peculiar importance from the fact that it 

 affects, to a relatively large amount, the zenith-distances of polar- 

 stars, and hence the deduced results for latitude. The error is 

 eliminated in the mean of a reflex and a direct observation 

 taken at the same time. James Renton, 



SUGAR-CANE EXPERIMENTS. 

 JN the fourth number of the West Indian Bulletin, recently 

 noticed in these columns, many pages were devoted to com- 

 munications to Dr. Morris, the Imperial Commissioner of 

 Agriculture, from Prof d'Albuquerque, the Island professor of 

 chemistry, and Mr. Bovell, the agricultural superintendent, in 

 which an elaborate plan was laid down for undertaking an 

 exhaustive investigation into the merits of several varieties of 

 sugar-canes. The very full details therein given should be con- 



NO. 163 1, VOL. 63] 



suited by any one desirous of mastering the significance of the 

 facts contained in the pamphlet now issued by the Commis- 

 sioner, giving a "summary of the results of the cultivation of 

 seedling and other canes at the experiment stations in Barbados 

 in IQCXJ." Prof. d'Albuquerque and Mr. Bovell have read a 

 paper on the subject before the Barbados Agricultural Society 

 on the results of the cultivation and yield of selected 

 seedling and other canes, and the pamphlet summarises the 

 essential facts. It is important to remember that the experi- 

 ments were throughout conducted on the ordinary system of 

 natural cultivation, the planters themselves undertaking to set 

 apart plots of their own plantations, so that the known and the 

 unknown grew side by side, no exceptional treatment being 

 recognised. In this way fairly typical results are obtained, and 

 the results for subsequent years will, therefore, be watched with 

 more than usual interest to see how the character of the season, 

 as well as the quality of the soil, may affect the various canes. 

 For the experiments seven stations were selected, representing 

 the typical soils and climatic conditions of Barbados. Five of 

 the stations were black soil, the other two red soil. At 

 nearly every station there were duplicate plots of each variety, 

 serving to show the variation to be expected with each variety 

 from one part of the field to another. The lowest station was at an 

 elevation of 100 feet above sea-level, the highest 910 feet, the 

 rainfall in the growth period ranging from 56 inches to 89 

 inches. Fifteen selected varieties of canes were tested on the 

 black soil estates, and ten of them on the red soil estates. For 

 each variety the highest and the lowest yield in tons per acre in 

 the black and the red soils respectively are given, and separate 

 tables for black and red soils show for each cane the number of 

 plots used for the investigation, the yield in tons per acre of 

 canes and also of tops ; the juice per cent, by mill ; pounds 

 per gallon of saccharose, of glucose and of solids not sugar ; the 

 quotient of purity of the normal juice ; the juice in gallons per 

 acre ; saccharose in pounds per acre ; and the sugar in tons per 

 acre, calculated according to a formula supplied by Mr. Douglas, 

 of the Diamond plantations, British Guiana. In the black soil 

 B. 147 heads the list with 3'Oi tons of sugar per acre, followed 

 by B. 347 with 2 90 tons, and B. 208 with 2*83 tons, while at 

 the bottom of the list stand D. 145 with i'82 tons, the Burke 

 with I '73 tons, and the Bourbon with only 0*47 ton per acre. 

 The White Transparent cane, which is cultivated in Barbados 

 on a larger scale than any other cane, and may therefore be 

 regarded as the standard for comparison, occupies a middle 

 place with a yield of 241 tons per acre. In the red soil B. 208 

 takes first place with 3 34 tons per acre, followed closely by 

 B. 156 with 3 32 tons and B. 147 with 3'3i tons, the lowest 

 being B. 347 with 2'I7 tons and B. 254 with 2'I4 tons. 



The mean results for both soils indicate B. 147 to be the 

 best all-round cane, its yield being 27 52 tons of canes per acre, 

 31 tons of sugar and 6291 lbs. of saccharose, B. 208 occupy- 

 ing second place with respectively 22 '5 S tons, 3*02 tons, and 

 5443 lbs., compared with the standard White Transparent re- 

 sults of 20 "49 ions, 2 '4 1 tons, and 4528 lbs. A further table 

 gives the results obtained on the three estates of " Dodds," 

 "Pine "and " Waterford " with B. 147 and the White Trans- 

 parent varieties, the means for the three estates giving B. 147 a 

 yield per acre of 6999 lbs. of saccharose and 370 tons of sugar, 

 while the White Transparent yielded 4527 lbs. of saccharose and 

 2'4i tons of sugar. It will thus be seen that the new seedling, 

 B. 147, is better than the standard by more than a ton of sugar 

 per acre. Looking to the individual and general results, the in- 

 vestigators consider there is a satisfactory degree of agree- 

 ment under a considerable variety of conditions of culture and 

 growth. B. 147 is regarded as the best all round seedling 

 variety as a plant cane in Barbados, B, 208 giving promise of 

 proving a good red soil plant and ratoon cane. Planters are ad- 

 vised to try, on a small scale, three or four of the varieties which 

 have done best in these experiments, so as to be able to secure 

 eventually the cane best suited to the nature of their particular 

 fields and their own methods of cultivation — features which 

 have, in their way, quite as much weight as the character of 

 the cane itself While B. 147 seems to be the most suitable 

 cane for particular soils in Barbados, D. 95 appears to be the 

 best for the different circumstances of Aniigua. A private letter 

 from Barbados, in which reference is made to the above ex- 

 periments, states that the officials of the Agricultural Depart- 

 ment seem determined on securing improved varie es o 

 that will suit each district, and will yield at least 50 per cent, 

 more sugar than those hitherto cultivated. 



