I lO 



NATURE 



\_Dec. 2, 1880 



the decrease was followed by the wave of high pressure whicli 

 preceded the Orissa famine. 



Hence it appears that widespread and severe famines are gMO-ally 

 accompanied or immediately preceded by -waves of high barometric 

 prtsstire. 



Means whereby future famines may possibly be foreseen. — If the 

 conclusions arrived at from the above comparisons of abnormal 

 barometric variations, sun-spots, and past famines be admitted, 

 it is clear that they at once present the means whereby future 

 famines may possibly be foreseen. The conclusions are 

 briefly : — 



1. That variations of the solar spotted area are succeeded many 

 months afterwards by corresponding abnormal barometric varia- 

 tions. 



2. That abnormal barometric variations in the tropics travel 

 at a very slow rate lound the eirth from west to east, arriving 

 at westerly stations several months before they reach more easterly 

 ones. 



3. That famines follow in the wake of waves of high barome- 

 tric pressure. 



Hence it follows that there are two methods by which early 

 intimation of the approach of those meteorological disturbances 

 which are attended Ijy famines may possibly be obtained — 



1. By regular observation of the solar spotted area, and early 

 reduction of the observation^, so as to obtain early information 

 of current changes going on in the sun. 



2. By barometric observations at stations differing widely in 

 longitude, and the early communication of the resitlts to stations 

 situated to the westward. 



With regard to the first of the;e methods it ;is sufficient to 

 state that the whole subject of solar observations is now being 

 investigated by a committee of scientific gentlemen in London, 

 and we may therefore hope that the all-important information 

 which solar observations are capable of affording will ere long 

 be at our disposal ; but witii regard to the second method, viz., 

 that of barometric observations at stations differing widely in 

 longitude, it is to be regretted that no observatories of long 

 standing situated in suitable localities to the west:vard cif Bom- 

 bay at present exist, except possibly at the very distant station 

 of Havanna in Cuba. The observatory at St. Helena appears 

 to have been closed in the year 1847, after working continuously 

 for about seven years. 



The most suitable localities for barometric observations for 

 the purpose in view are insular stations far removed from the 

 disturbing influences of the large continents and near the equator, 

 such as the Seychelles, St. Helena, and Ascension, but these 

 appear to be at present unoccupied by permanent observatories, 

 while the wide expanse of the Pacific, which is probably the 

 most suitable portion of the earth's surface for investigations of 

 this kind, appears to be entirely unrepresented by any fixed 

 observatory on any of its numerous island*, such as the Gala- 

 pagos, Sandwich, and Fiji Islands. An observatory has how- 

 ever lately been established at Zanzibar on the East Coast of 

 Africa, from wliich very valuable observations may be expected 

 if it should continue at work for any great length of time ; and 

 another has, I believe, been started at Aden : but as these 

 stations are both situated on the borders of extensive continents, 

 they are not so suitably located as the stations previously 

 mentioned. 



It would therefore be necessary, in order to utilise to the fullest 

 extent the second method of foreseeing the approach of a 

 meteorological disturbance of the kind which would probably be 

 attended by famine, that special arrangements should be made 

 for the registration of the needful observations at some, if not 

 all, of the stations that have been referred to ; and that the 

 information thus afforded should be rapidly communicated from 

 the more westerly to the more easterly stations. 



F. Chambers, 

 Meteorological Reporter for 



Bombay, September 4 Western India 



Postscript. — In order to determine numerically the intervals 

 of time at which the barometric variations of one station have 

 lagged behind those of another, and behind corresponding minor 

 variations of the sun spots, the times at which the continuous 

 curves cross the dotted ones have been marked off by the graphic 

 method for corresponding crossing points of the different curves, 

 giving the first set of times and intervals in each of the following 

 tables. The same thing has been done with regard to the times 

 at which the continuous curves cross the respective zero lines. 



giving the second set of times and intervals in each of the tables. 

 As the average pressures for Batavia and Bombay have not been 

 calculated from the observations of the same years, and as the 

 zero line of the Batavia curve is on this account relatively dis- 

 placed by '004 of an inch in the upward direction, a new zero 

 line has been drawn so as to make the times at which the con- 

 tinuous curve crosses the zero line comparable with those for 

 Bombay. The approximate longitudes of the stations and their 

 differences are also given in the tables. 



