376 RECORD OP SCIENCE FOR 1887 AND 1888. 



{^^}) Jylatt/orfl. — "A inactical ^uulc to the climates and weather in 

 India, Oeylon, and Jiuiinah, and the stoniis of Indian seas," by II. y. 

 Bhmford. This vohune, pnblished in 1889, marks the retirement of Mr. 

 Blauford as meteorological reporter to the Government of India, in 

 which position, since 1875, he has undoubtedly accomplished the great- 

 est works ever undertaken in meteorology. Wilde's Repertorium and 

 Annalen, Mascart's four annual volumes of observations and memoirs, 

 Neumayer's annual Ergebnisse and Archiv, are the principal works to be 

 named in comparison with Blanford's annual reports and Indian mem- 

 oirs. 



In the present work Blanford devotes only a moderate portion to 

 storms, and summarises the results that have been attained in India by 

 Piddington, Willson, Pedlar, Meldruin, and especially the great work 

 of John Eliot. 



So far as could be seen from the scattered records available up to the 

 end of 1876, cyclones are most frequent in the Bay of Bengal in May 

 and October, and least frequent, being almost unknown, in February 

 and very rare in July ; but with the more perfect data furnished by the 

 Government weather office Mr. Eliot has compiled a list of cyclones for 

 the ten years 1877-8(3, which entirely alters this supposed annual dis- 

 tribution, and makes the distribution almost uniform from June to No- 

 vember. The Indian storms generally move toward northwest or west- 

 northwest, and with very few exceptions the direction is toward some 

 point in the northwest quadrant. 



The incurvature of the winds for storms in the northern part of the 

 Bay of Bengal is about 35 degrees, with a slight increase as we go 

 southward. The barometer does not fall for the approaching cyclone 

 until long after the wind and skies show that the center of the storm 

 is close at hand and only when the winds become strong and squally, 

 nor does it fall rapidly until the winds have increased to hurricane 

 force and the center is distant less than 50 miles. The most important 

 indications of the approach of a storm are therefore to be found in the 

 observation of the sky and clouds. 



The storms of the Arabian Sea, on the west coast of India, have been 

 studied by Mr. F. Chambers. The storms that cross India as they move 

 westward deteriorate greatly before they reach the Arabian Sea, but 

 then at once begin to increase. The stormy months are April, May, 

 June, and November, but the records are rather fragmentary, and the 

 results will probably be changed by future study. The system of storm 

 signals for the Bay of Bengal is under the control of the Calcutta oflice, 

 and consists of a simple warning or danger signal at all ports except 

 Calcutta, where a much more elaborate system is used. There is an 

 independent system of storm warnings for the Arabian sea-coast by the 

 office at Bombay. 



(IG) Davis. — "Whirlwinds, cyclones, and tornadoes," by William M. 

 Davis. This little book is the result of a series of lectures for the 



