490 



METEOROLOGY. 



object be weather prediction, a detailed termi- 

 nology is necessary, and for this purpose, Mr. 

 "W. 0. Ley's classification is regarded as unsur- 

 passed. One of the principal objects is the 

 determination of the direction of the wind in 

 the higher regions of the atmosphere, and for 

 this it is not necessary to distinguish so many 

 forms; but we must be sure that these forms 

 are, generally speaking, everywhere the same ; 

 and their mean heights must be severally de- 

 termined by direct measurements. 



M. Teisserenc de Bort concludes, from his 

 studies of the distribution of cloudiness over 

 the earth, that there is a marked tendency of 

 cloudiness throughout the year to arrange it- 

 self in zones parallel to the equator. Aside 

 from perturbations complicating the matter, 

 there may be traced a maximum of cloudiness 

 in the region of the equator; two bands of 

 light cloudiness extending from 15 to 35 of 

 latitude north and south, and two zones of 

 greater cloudiness between 45 and 60, be- 

 yond which, so far as we can judge from ap- 

 pearances in the northern hemisphere, the sky 

 seems to become clearer toward the poles. 

 These zones have a marked tendency to follow 

 the sun in its change of declination, shifting 

 up toward the north in the spring and toward 

 the south in the fall. Further comparisons in- 

 dicate that the zones of clear sky correspond 

 with the regions of high pressures on either side 

 of the equator, and originate, on one hand, the 

 trade-winds, and on the other hand, the west 

 winds, which are dominant near the temperate 

 regions of both hemispheres. The zones of 

 more cloudy sky extend above the regions of 

 low pressure ; that is, near the equator and near 

 the 60th degrees of latitude. The study of the 

 winds shows that the air, at the surface of the 

 soil, diverges from the zones of high pressure 

 beyond the tropics, and blows toward the 

 regions of low pressure, or toward the equa- 

 torial regions, and toward the low pressures 

 near the 60th degrees. Hence, probably, that 

 the winds near the centers of divergence have 

 a descending component, and in the regions 

 toward which they blow, an ascending com- 

 ponent, consequently, other things being equal, 

 the cloudiness is light when the wind has a 

 vertical component directed downward, and 

 heavy when the direction is upward. This 

 agrees with the fact that an ascending mass of 

 air cools in expanding, and thus tends to pro- 

 duce a condensation of aqueous vapor, while 

 the contrary takes place with a descending 

 mass. The distribution of cloudiness is a di- 

 rect consequence of the course of the winds. 



The attention of certain observers has been 

 given since 1885 to " the strange effect of 

 bright, silvery-lighted clouds," which have re- 

 mained visible at times in the northwest sky 

 after sunset till late in the evening. While 

 presenting varieties in structure, they are de- 

 scribed as usually exhibiting an opaque, pearly 

 luster, with definite outline. They were first 

 noticed by Mr. D. J. Rowan, of county Dublin, 



Ireland, some two or three years previous to 

 1886, who regards them as sub-auroral and 

 self-luminous, and proposes for them the name 

 nuberculce boreales. They are also regarded as 

 auroral by Prof. Piazzi Smith, who has made 

 spectroscopic observations of them, although 

 he admits that their spectrum is that of twi- 

 light. Mr. J. W. Backhouse, of Sunderland, 

 England, on the other hand, is of the opinion 

 that they derive their light from direct solar 

 illumination, and points to the fact that they 

 have so far been seen only during a short 

 period before and after the summer solstice, 

 the recorded dates of observations ranging 

 from May 28 to August 12. 



Important researches on the motion and 

 height of clouds have been carried on at the 

 University of Upsala, Sweden, under the di- 

 rection of Prof. Hildebrandsson, with the assist- 

 ance of Messrs. Ekholm and Hagerstrom. A 

 peculiar form of altazimuth, which was origi- 

 nally designed for measuring the parallax of 

 the aurora borealis, is used to take the angular 

 measurements. The greatest height of cloud 

 which has yet been measured is 43,800 feet ; 

 the highest velocity is 112 miles an hour with 

 a cloud at 28,000 feet. The clouds are not 

 distributed promiscuously at all heights in the 

 air, but have a tendency to form at three definite 

 levels. The mean summer altitude of the three 

 levels has been found to be: low clouds stratus, 

 cumulus, cumulo-nimbus, from 2,000 to 6,000 

 feet; middle clouds strato-cirrus and cumulo- 

 cirrus, from 12,000 to 15,000 feet; and high 

 clouds cirrus, cirro-stratus, and cirro-cumulus, 

 from 20,000 to 27,000 feet. The velocity ob- 

 servations confirm the results that have been ob- 

 tained from mountain stations, viz., that, though 

 the general travel of the middle and higher 

 clouds is much greater than that of the surface 

 winds, the diurnal variation of speed at those 

 levels is the reverse of what occurs near the 

 ground. The greatest velocity on the earth's 

 surface is at about 2 o'clock p. M., whereas the 

 lowest rate of the upper currents is at about 

 mid-day. The mean height of all varieties of 

 clouds rises in the course of the day, and is 

 greater at between 6 and 8 o'clock in the 

 evening than either in the early morning or at 

 mid-day. 



Rainfall. The hypothesis suggested by Mr. 

 Norman Lockyer, about 1873, that certain 

 periodical variations of the rainfall at Madras, 

 India, are connected with periods of maxima 

 and minima in the frequency of sunspots, has 

 been re-examined by Mr. Henry F. Blanford 

 in the light of the accumulated data, applying 

 especially to the Carnatic of the last twenty- 

 two years. Comparative tests of the rainfall 

 and of sunspot frequency, year by year, have 

 satisfied the author that, apart from the ap- 

 proximate identity of its period, the oscillation 

 of the rainfall is very different in character 

 from that of the sunspot frequency. Compari- 

 sons with the data of other provinces also 

 showed that the apparent periodical variation 



