424 RECORD OF SCIENCE FOR 1887 AND 1888. 



Such storms geuerally show themselves as well uucler way on the 3 p. m. 

 map. The geographical distribution of frequency of this class of 

 storms for the United States was illustrated by me in the map for Au- 

 gust, 1874, published in the Signal Service Monthly Weather Eeview 

 for that month. 



The paths pursued by these thunder storms trend eastward or north- 

 eastward, as in Europe, see Mohn and Hildebrandsson. They may 

 be classed as (1) those that attend a very low pressure, and (2) those 

 that start on the advancing edge of areas of cool, dry air. or high barom- 

 eter, and (3) those that occur within and on the rearward side of high 

 areas. Quite a number of them may exist simultaneously, for instance, 

 along the whole eastern slope of the Appalachian range, all moving in 

 nearly parallel Hues northeastward until they die out in the earlj" morn- 

 ing hours, generally before they reach the Atlantic shore, but sometimes 

 extending a little wa3\s out to sea. 



(3) Light general rains occurring between 1 and 8 A. m., and which are 

 apparently caused by radiation of heat from a layer of moist air some 

 distance abovethe earth. Frequently when the sky is perfectly clear 

 at 11 P. M. there prevails over a special region a general strong southerly 

 current feediug a storm center to the northward, but not as yet followed 

 by any condensation in this special region. It would seem that the 

 nocturnal radiation of heat from the moist and hazy sky now comes in 

 powerfully to supplement the dynamic cooling, which latter process de- 

 pends on ascension, and therefore, of course, is less active during the 

 night than during sunshine; accordingly haze and cloud form after 11 

 p. M. and light general rains or snow fall in the early morning hours, 

 and are reported on the 7 A. M. map ; these are liable to be stopped by 

 the sun's heat as soon as that is sufficient to counteract the radiation, 

 but when this heat first falls on the cloud top it stimulates the convec- 

 tion and the precepitation. The only means of anticipating the occur- 

 rence of this class of rains consists in a close observance of all the ante- 

 cedent phenomena in order to estimate the probable relative tempera- 

 ture and moisture of the air overhead ; the only observations bearing 

 on this subject, as shown by the 11 p. m. ma]i, are a slight tendency to 

 haze, the formation and disappearance of loose clouds, a slight check in 

 the temperature, and especially the reports of halos around the moou, 

 to which, according to Montigny, it is important to add the observation 

 of the scintillation of the stars. 



(4) Extended heavy rains. These generally attend either areas of 

 low pressure, where great masses of air apparently have slow ascend- 

 ing and vorticose movements, or prevailing winds tiiat carry great 

 masses of moist air up mountain pr plateau slopes to higher levels; 

 these rains are theretore due principally to dynamic cooling, and are 

 well illustrated by easterly storms of the Atlantic coast, the northerly 

 storms of the Central American and Mexican coasts, and the southerly 

 windsof the Gulf States. Their prediction is implied in the prediction 

 of a general storm as before treated of. 



