282 SCIENTIFIC RECORD FOR 1884. 



Jmic 2. 1883, wherein he gives reasons for believing that the prospects 

 lor sullicient rain in India and Bengal during the rest of the season 

 were wlioUy favorable. This and subsequent predictions in July and 

 August were based upon the knowledge of snowfall in the Himalayas. 

 {Nature, xxix, p. 77.) 



95. C L.Madsen, in the interest of meteorology and European weather 

 prediction, urges the importance of the proposed transatlantic cable, 

 whose route will be through Thurso, the Faroes, Iceland, Julianshaab 

 in Greenland, the Straits of Belle Isle, and Quebec. The German news- 

 papers announce that this cable will be laid by a company already or- 

 ganized ; but although this is probably an error, yet there seems no 

 doubt but that such a cable will be a profitable investment from a busi- 

 ness point of view, and will be of the highest importance in the predic- 

 tion of storms, not only for the use of Europe, but especially for the use 

 of navigators about to sail westward from those ports. (Z>. M. Z., i, 

 p. 410.) 



96. The British Meteorological Office has entered into an arrangement 

 with the Signal Office by which telegraphic warning is sent to British 

 ports of the existence of any storm in the middle of the IsTorth Atlantic 

 into which English vessels are likely to enter unawares. 



97. R. Abercromby, W. Marriott, Lieut. H. H. C. Dunwoody, and Prof. 

 H H. Hildebrandsson, of Stockholm, have all recently published collec- 

 tions of popular sayings in reference to the weather. The latter espe- 

 cially expresses the following conclusions : (1) Any relation between the 

 weather and the days of the week, or holidays, or the x^ositions of the 

 heavenly bodies, or the phases of the moon is absurd ; (2) any predic- 

 tion of the temperature of any year or month is impossible in the pres- 

 ent state of our knowledge, and popular signs looking to such predic- 

 tions are without value; (3) in general, it is desirable for vegetation and 

 harvest that the seasons should have their normal temperature, rainfall, 

 &c., and that is about what is attempted by all jiopular rules ; (4) there 

 are a number of rules for the prediction of the weather a short time in 

 advance, some of which are quite safe. {Z. 0. G. M., xix, p. 311.) 



98. R. Abercromby presents a summary of prevailing weather types 

 for use in predictions in Western Europe. {Nature, xxviii, p. 330.) 



99. E. Gelcich communicates the rules promulgated by P. Jauva, di- 

 rector of the Observatory at Manila, for the prediction of hurricanes. 

 During the years 1877 to 1880 Jauva has successfully predicted forty 

 storms. The following are his rules for Manila : 



1. If the barometer at the time of the afternoon minimum, namely, 

 between 3.15 and 3.30, falls to 7o5""", or if it exceeds 757 at the time 

 of the morning maximum, namely, from 9.10 to.9.20, then one can defi- 

 nitely conclude that a cyclone prevails within a few hundred miles. 



2. If at the time when the barometer should rise, namely, from 4 to 9 

 A. M., or from 4 to 9 p. m., it is observed to fall, then the cyclone is mov- 



