56 Mr Lyall on the Weather in Madagascar. 



GENERAL OBSERVATIONS. 



1. TANANARIVOU, the capital of Madagascar, is situated in 18 56' 20" 

 S. L., and, I conjecture, in about 47 E. L. From barometric obser- 

 vations, I reckon its elevation to be nearly five thousand feet above 

 the level of the sea ; and its highest pinnacle, called Ambouin Sim- 

 boun, about seven hundred and fifty feet above the level of the greatest 

 part of the surrounding plain. I am about to make extensive and 

 more accurate observations respecting some of these points, which I 

 shall not fail, in due time, to make public. 



2. In consequence of the peculiar situation of Tananarivou, and 

 especially of its great elevation, a series of well-conducted and well- 

 recorded meteorological observations must be of the highest interest. 

 By the acquisition of additional instruments, and greater practice, I 

 trust to render every month's report more detailed and more interesting 

 than another, until the climate here is sufficiently known. 



3. The observations have been made every morning at six o'clock, 

 because this is the only hour on which I could count for regularity : 

 therefore, the day commences at six o'clock, A. M., and ends at six, 

 A. M., of the succeeding day. 



4. The sympiesometer used is Adie's, No. 497. 



5. In consequence of an accident having happened to one of 

 Newman's mountain barometers (an excellent instrument), I have 

 been compelled to make the foregoing observations with Jones's 

 mountain barometer, which is constantly suspended. In order to 

 have the means of making comparative observations, however, I my- 

 self filled the tube of Newman's barometer, which, though the starting 

 point be somewhat different, acts upon the same general principles and 

 in the same manner as Jones's. The two instruments work together. 



6. Rutherford's register thermometer is used for the maximum and 

 minimum. 



7. In all my observations with the sympiesometer, after carefully 

 adjusting thejtfewr de lis, I find it necessary to attribute 10, 12, 14, 

 16, or even more degrees to the mere effect of temperature, between 

 the hours of one or two o'clock and four o'clock, p. M. ; otherwise I 

 should be constantly predicting rain. 



8. Daniell's hygrometer is used. 



9. I have given a statement of facts, without attempting to draw 

 conclusions. Time does not permit such inquiries ; besides, professed 

 meteorologists will do this much better than I could : therefore, copies 

 of this table, and of that of all future tables, shall be forwarded to my 

 friends, Mr. Dalton, of Manchester, and Mr. Daniell, of London. 



ROBERT LYALL. 



