378 RECORD OF SCIENCE FOR 1887 AND 1888. 



(g) the immediate antecedent is heavy rainfall, concentrated over a 

 portion of the bay ; (/^) accompanied by a strong indraft, which is most 

 marked from the Indian Ocean at the entrance of the bay; {i) this in- 

 draft from the Indian Ocean gives rise to strong winds and heavy rains 

 at the stations on the south and west coast of Ceylon." 



Of the preceding items {a) (b) (c) are invariable and necessary ante- 

 cedents, bnt the source of the energy is the item {g), condensation of 

 vapor and precipitation of rain, or, as stated in the Backergunge report, 

 the primary cause of cyclone formation is the production and ascent of 

 a large quantity of vapor, which is condensed with the liberation of its 

 latent heat over the place of its production instead of being carried 

 away to some distant region. 



This independent confirmation of the views for which Espy lived and 

 died, as well as the numerous other generalizations not quoted by us, 

 after having been clearly apprehended by their author, were found re- 

 peated in other storms, and were confirmed in his subsequent memoir 

 "On the cyclonic storms of November and December, 1886, in the Bay 

 of Bengal." Eliot states that there is a marked difference in the char- 

 acteristics of the storms of one year and of another, which is explicable 

 on the same hypothesis as that which explains the variations in the 

 sovithwest monsoon rain-fall. These variations are apparently periodic 

 in Bengal, and when the Bengal branch of the monsoon current be- 

 comes strong the Bombay branch becomes weak. 



Pending the appointment of a successor to Mr. Blanford, Eliot, as 

 officiating reporter, has published the report of 1887 on the Meteorol- 

 ogy of India. On page 209 he refers to certain storms in Bengal known 

 as northwesters, which are occasionally as destructive as the tornadoes 

 of the United States, but his description does not enable us to conclude 

 as to whether they were of the nature of twisting tornadoes, or the 

 straight-line Derecho, described by Hinrichs, in Iowa. 



On page 251 Eliot says the great majority of cyclonic rain-storms 

 march across the Bengal coast in the direction of the belt of lowest 

 pressure at the time of their formation. As the chief characteristic of 

 such a barometric trough is light and variable winds, it will be seen 

 that this principle virtually coincides with the rule of cyclonic storms 

 in the Bay of Bengal, which marcli in the direction of least relative air 

 motion immediately antecedent tX> the formation of the cyclone. 



On page 271 Eliot states that " the persistency of the pressure anomo- 

 lies (for weeks and months) is almost certainly due to the fact that an 

 abnormal variation of pressure in a moving mass of air necessarily 

 gives rise to or accompanies the modification of its motion and in con- 

 sequence of well-known properties of fluid motion this changed or 

 modified air motion tends to perpetuate the pressure variation which 

 gave rise to it." 



If I correctly understand this sentence I should apply it to the flow 

 of air over an obstacle where the change of motion produces a change 



