METEOROLOGY. (STORMS, RAINFALL.) 



535 



dicated between the numbers of dust particles 

 and the cyclones and anti-cyclones over north- 

 western Europe at the time. The observations 

 also indicate that the dust particles may vary 

 enormously during mist or fog, without any dif- 

 ference in the apparent density of the fog. 



Atmospheric dust is divided by Dr. W. Mar- 

 cet into organic and inorganic. The dust scat- 

 tered everywhere in the atmosphere, which is 

 lighted up in a sunbeam or a ray from an elec- 

 tric lamp is organic. It is seen to consist of 

 countless motes, rising, falling, or gyrating, al- 

 though it is impossible to follow any of them 

 with the eye for longer than a fraction of a sec- 

 ond. It is difficult to say how much of the dust 

 present in the air may become a source of dis- 

 ease and how much may be innocuous. Many 

 of the motes belong to the class of micro-organ- 

 isms which are frequently the means of spread- 

 ing infectious diseases. 



Storms. Four classes of summer thunder 

 storms are described by W. M. Davis in the Ob- 

 servations of the New England Meteorological 

 Society. Those of the first class are of local 

 origin, on quiet, hot, anti-cyclonic days, the ac- 

 tion of which suggests that they are merely over- 

 grown convectional movements. They are less 

 common than those of the other classes. A sec- 

 ond class consists of those storms which spring 

 up in the warm southerly winds southwest of a 

 cyclonic center, whose convectional overturning 

 is therefore due in part to imported heat. They 

 are also relatively rare. In the third class are 

 included the largest of our summer thunder 

 storms, which seem to be produced where the 

 warm southerly winds are most nearly contrasted 

 with the cooler westerlies that follow them ; the 

 storm forms along the line between the two, and 

 advances obliquely, across country. The fourth 

 class of thunder 'storms contains those which 

 arise in the westerly winds southwest of a cy- 

 clonic center, and whose convectional overturn- 

 ing is due as much to the importation of cool 

 air aloft as to its warming on the ground below. 

 They are of the same kind as the snow squalls 

 of late winter, except that their violence is 

 greater. The storms of the first and second 

 classes are not followed by cooler weather, but 

 those of the third and fourth classes generally 

 accompany a moderate fall of temperature, 

 brought by westerly winds. The more local 

 storms of the first, second, and fourth classes 

 are of moderate size, a few miles wide and from 

 five to thirty miles long. The larger storms of 

 the third class are hardly more than ten or 

 twelve miles wide, but may be one or two hun- 

 dred miles long, or even longer. These move 

 obliquely broadside with the wind rushing out 

 in the front. The longer storm fronts may not 

 be truly continuous, and yet the different masses 

 advance so uniformly that they may all properly 

 enough be associated as parts of a whole. The 

 longer fronts generally trend northeast and 

 southwest, while the general course of the storm 

 is a little north of east. So uniform is this ad- 

 vance, both in direction and velocity, that it is 

 difficult to accept the common belief in the con- 

 trol exerted by rivers and valleys over the path 

 of thunder storms. Tornadoes are developed in 

 these thunder storms only on rare occasions. 

 The occasional hail-falls of summer are always 



associated with thunder storms. They are not 

 rare, and yet are not common enough to cause 

 much destruction. Like tornadoes, they seem 

 to be local developments of exceptional strength 

 within the body of a thunder storm ; and, like 

 tornadoes, the belts of country over which they 

 are distributed trend, as far as has been made 

 out, eastward or northeastward. 



Five years' studies of thunder storms, from 

 1882 to 1886, made on the Hungarian plain by 

 M. Hegzfoky, show that on the days on which 

 thunder was observed, forming 16*4 per cent, of 

 all days from April to September, the air press- 

 ure sank about 2 millimetres under the normal, 

 morning and evening. The less the pressure, 

 the greater the probability of a thunder storm. 

 The temperature and the moisture and cloudi- 

 ness were in excess. The wind blew about 

 mid-day more softly, and in the evening more 

 strongly than usual. It went round, as a rule, 

 from the southeast by the south to the west 

 and northwest. The clouds came oftener than 

 usual from the southeast and southwest quad- 

 rants ; so that the center was usually north of 

 the station. Nearly half of the season's rain- 

 fall was on days of thunder storm. The first 

 thunder of the thunder storm most often occurred 

 between 2 o'clock and 5 o'clock, p. M. Toward 

 the end ef the season the storms tended to come 

 later in the day. When the pressure fell below 

 the mean of the season, the thunder storms lasted 

 longer than when it was above the mean. After 

 the first thunder the meteorological elements 

 were usually subject to great changes, most 

 marked as the storm neared the zenith ; rain 

 fell, wind rose and altered quickly in direction, 

 temperature and vapor pressure fell, and relative 

 humidity, cloud, and pressure increased. As the 

 storm withdrew there was a return to the normal. 



The development of storms is attributed by 

 E. B. Garriott, in the " American Meteorological 

 Journal," to an excess of heat received from the 

 earth's surface by radiation, and their progressive 

 movement to the precipitation of aqueous vapor 

 at a considerable height ; while the direction in 

 which they move is affected by several conditions, 

 including the disposition of cold, dry air found 

 in areas of high pressure. The areas of low 

 pressure in endeavoring to advance eastward 

 seem to avoid the heavier bodies of cold, dry air, 

 and move toward the point where there is the 

 least resistance to their advance, which would 

 naturally be toward warm, moist regions. Thus, 

 of the storms of North America, a large majority 

 originate over the plateau region in the lee of 

 the Pacific coast ranges of mountains, and ad^ 

 vance toward the regions of greatest moisture 

 which embrace the Great Lakes, the Gulf of 

 Mexico, and the valleys of the principal rivers. , 



Dr. Doberck, of the Hong-Kong Observatory, 

 has expressed the opinion, from comparisons act- 

 ually made there, that the indications of spectro- 

 scopic rain-band observations frequently foretell 

 thunder storms which could not otherwise have 

 been forecast from local observations. 



Rainfall. A distinction is made by M. H. 

 Fischer in his studies upon the equatorial limits 

 of snow between regular falls that is, a certain 

 average amount in each winter and those falls 

 which are only occasional. The following table 

 exhibits some of the author's results : 



