344 



NATURE 



[September 14, ign 



CYCLONES AND SUGAR-CANES. 

 The Sugar Industry of Mauritius : a Study in Correla- 

 tion. Including a Scheme of Insurance of the 

 Cane Crop against damage caused by Cyclones. 

 By A. Walter. Pp. xvi + 228. (London: A. L. 

 Humphreys, 1910.) Price 12s. 6d. net. 



MAURITIUS, situated just within the tropics and 

 on an edge of the southern anticyclonic system, 

 is in certain circumstances visited by tropical cyclones. 

 Occasionally great damage to the growing sugar 

 crops is done by such visitations, and though ex- 

 tremely violent storms are exceptional, yet the fact 

 that they may and do occur makes the cyclone season 

 one of dread and anxiety in the island. In the year 

 1892, for example, about one-half of the total crop 

 was destroyed, to say nothing of damage to build- 

 ings, and other losses. True, this storm was one of 

 quite unusual severity ; but fear of a similar calamity- 

 tends, nevertheless, to have a paralysing effect upon 

 the sugar industry. " Behind every attempt at im- 

 provements and every fresh outlay of capital, hovers 

 the spectre of the 1892 disaster." 



In these circumstances the question of insurance 

 becomes one of primary importance. With a view to 

 obtaining some rational basis for proposals of insur- 

 ance the investigation described in the volume before 

 us was undertaken : it is an attempt to determine 

 quantitatively the relative effects of rainfall, tempera- 

 ture, and cyclonic wind-velocity upon the amount of 

 the sugar-crop. 



Obviously the problem is a complicated one. The 

 author attacks it by a study of the climatic statistics 

 for a number of years, in correlation with the crop 

 harvested in each of those years. By successive 

 approximations, after allowing for exceptional in- 

 fluences, it has been found possible to evaluate the 

 effects of the several climatic factors upon the crop. 



Taking first the question of rain, the changes in 

 the amount of crop were found to follow in a general 

 sense those of the rainfall, but the ratio crop/rainfall 

 varied considerably from year to year. There is, 

 however, not merely the total quantity of rain to be 

 considered ; the number of rainy days was soon seen to 

 be as important as the total rainfall. From an agricul- 

 tural point of view a month in which five inches of 

 rainfall was distributed over twenty days would be a 

 wetter month than if the whole had fallen in one 

 day. Hence in order to obtain the combined effect of 

 quantity and chronological distribution, the idea of 

 "degree of wetness" for each month was introduced 

 by means of the expression R.i'/ f > where R is the 

 total rainfall for the month, t the number of days, 

 and t l the number of rainy days. But further, in 

 excessive falls a part of the rain is non-effective, and 

 it was necessary to eliminate this portion, partially 

 at least. The method adopted depended upon the 

 consideration that if it were possible to determine the 

 depth of soil to which a given fall penetrated, the 

 amount which passed beyond the limit of the cane 

 roots might be rejected. Increase of temperature in 

 the soil at certain depths was taken as a guide to 

 the points beyond which the fall penetrated. 

 NO. 2185, VOL. 87] 



The effect of high wind velocity on the cane is 

 regarded as falling under one or all of the following 

 heads :— (a) Tearing of the leaves ; (b) bending of the 

 stem about its weakest part; and (c) loosening and 

 rupture of the roots. The magnitude of the effect (a) 

 depends upon the pressure P of the wind, the duration 

 h of the high velocity, and the epoch t in the plant's 

 growth at which it occurs, since the riper the cane 

 the greater the damage; effects (6) and (c) depend 

 also on the amount of change 9 in the direction of 

 the wind, the greatest damage being caused when 

 the cane is bent in directions differing by 180 . 

 Whence the total effect on the crop is deduced to be 

 expressible by the formula icP 2 /i(2 + sin J0) sin if. 

 The value of the constant k is obtained from the crop 

 results given in three cyclone years, after eliminating 

 the rainfall and temperature effects. This formula 

 is suggested as a basis for determining the loss due 

 to a cyclone, the values of P, h, and 9 being taken 

 from the official observatory records. The practical 

 application for purposes of insurance is explained in 

 detail. 



The work is an ingenious and interesting statistical 

 study which may prove to have practical value. A 

 prediction of the crop yield for 1908, based on the 

 methods explained, turned out to be correct within 

 3 per cent. There seems no reason why, if circum- 

 stances demanded it, the methods should not be 

 applicable to other crops than sugar. 



c. s. 



IRRIGATION ON THE INDUS. 

 Punjab Rivers and Works. By E. S. Bellasis. Pp. 

 vii + 65 + 47 figs. (Allahabad : printed at the Pioneer 

 Press, 191 1.) 



THE author presents this work as a treatise setting 

 forth the general rules and principles on which 

 the inundation canals and flood embankments of the 

 Punjab are designed and maintained, since no such 

 work has hitherto existed. This we may take as 

 equivalent to saying that the regulations and practice 

 of the Punjab Irrigation Service have not yet been 

 systematised and brought into book form, since the 

 present volume deals with this aspect of the subject, 

 and does not, unfortunately, cover the wider field of 

 the hydrography of the Indus in its bearing on the 

 utilisation of its water. 



While the first chapter deals very briefly with the 

 rivers of the Punjab and their physical characteristics, 

 the next three discuss in considerable detail the inunda- 

 tion or flood canals, the flood embankments, and the 

 river-training works ; the last chapter and seven appen- 

 dices deal with the ordinary procedure, and give 

 certain regulations, orders, and specifications as issued 

 for the work of the Service. The information relating 

 to canals deals with their maintenance, silt deposit, 

 erosion, their section, alignment, and the various 

 modifications which have to be adopted in dealing with 

 the water of a river which has a high seasonal flood 

 and carries a heavy silt load. These characteristics 

 also necessitate flood embankments to control the 



