Chase] -^50 [May a and 16, 



In a few instances a line of rainfall with double curvature has been 

 strikingly marked, extending almost from the summit of an atmospheric 

 crest to the foot of an adjacent valley, thus exemplifying the accuracy of 

 Ferrel's deductions, and encouraging the belief that close and systematic 

 study of the tri-daily maps, under the guidance of those deductions, will 

 render the future success of our Signal Service Bureau even more mar- 

 velous than the past. 



Some stations seem to be exposed to peculiar local anti-cyclonic influences. 

 Although the number of stations is insufficient for an entirely satisfactory 

 study of such influences, and although I have given no special attention 

 to their investigation, having merely noted a few of the anomalies which 

 seemed most striking, I will venture to suggest a careful tabulation and 

 examination of reports from Shreveport, Lynchburg, Denver, Cheyenne, 

 Pittsburg, and St. Louis, with a view to subjecting any exceptional indi- 

 cations, which they may furnish, to a rigid scrutiny. I have sometimes 

 been inclined to attribute irregularities, which seemed to be of a system- 

 atic character, to imperfections in the instruments of observation, but 

 the careful comparisons to which they are subjected before leaving the 

 AVashington office, seem to make such a hypothesis less probable than 

 the one which looks to local perturbations, originating in peculiarities of 

 physical position. 



The tendency to parallelism of atmospheric currents, both in vertical 

 and horizontal planes, seems to be indicated in a large majority of the 

 maps. The isolation of opposite horizontal currents appears, however, 

 to be more complete than that of the vertical currents. So generally is 

 this the case that I doubt if there is ever any considerable blending of 

 upper and lower strata of air, except when a partial vacuum has been 

 brought about by great condensation of vapor through the whole height 

 of contiguous polar and equatorial currents, which are moving in oppo- 

 site directions in the same stratum. If my belief is well founded, the 

 cyclonism or anti-cyclonism of storms at their beginning should depend 

 mainly on the relative position of the mixing currents. As a general 

 rule, if the condensing current is east of the vapor-saturated current or 

 between the vapor current and the equator, the storm should be anti- 

 cyclonic, unless, and until, local precipitation has been copious enough 

 to reverse the normal direction of the blending winds. If the relative 

 position of the vapor-bearing and cooling currents is reversed, the storm 

 should become more speedily and increasingly cyclonic, the precipitation 

 should be more rapid, the winds more violent and tempestuous, the con- 

 ditions in other respects, such as pressure, vapor-saturation, temperature, 

 and velocity of wind at the outbreak of the storms, being the same. 

 These views seem to be confirmed by the facts which I have adduced in 

 previous papers, as well as by the special examination of which I am 

 now treating, but for that very reason I prefer to submit them, by simple 

 statement, for examination and test by others, who are either wholly free 

 from any prepossession, or whose bias is different from my own; my wish 



