METEOROLOGY. 317 



drops all contribute to the shade?] So long as such clouds retain their 

 sharp upper limit and their general spherical shape, no rain falls from 

 them. Suddenly the summit has a softer appearance, extends itself side- 

 ways in cirrus-like threads, and immediately the rain plunges down out of 

 the cloud. The electric tension which hindered the conjunction of the 

 water particles that form the cloud, so long as they retained their spher- 

 ical form, is suddenly diminished in the upper part of the cloud as soon as 

 the little s])heres freeze into ice needles or spiculte, from whose points 

 and corners the electricity immediately escapes. [The easy escape of 

 electricity from snow-crj^stals is a matter which should easily be con- 

 firmed by observation ; it agrees with the high potential observed dur- 

 ing snow-storms.] The minute bodies having lost their electricity, unite 

 and absorb in their rapid descent the smaller particles of water with 

 which they come in contact. [Do the spicula3 and snow-flakes descend 

 rapidly enough to overtake particles of water?] The falling rain, and 

 perhaps still more the sudden lightning, when such occurs, further con- 

 tribute to draw the electricity from the cloud or lower its electric i)oten- 

 tial ; the rain continues until the whole or nearly the whole lovfer part 

 of the cloud disappears. In those cases in which not only in the lower 

 but also in the upper strata of the atmosphere only slight movements 

 prevail, the ice-cloud remaining at the summit becomes a true cirrus, 

 whose curled and twisted form is probably a consequence of temporary 

 irregularities in the pressure of the air due to the condensation and 

 formation of ice. A cirrus which is formed in this way remains some- 

 times more than twenty-four hours almost perfectly without motion in 

 the sky, but more frequently it moves slowly into regions into which 

 the rain-shower out of which it originated has not been visible. If the 

 upper and lower parts of this cloud do not move at the same rai)idity 

 it will be drawn out into the long cirro-filum cloud, and this probably 

 often occurs when the upper part of the original cumulus has penetrated 

 some cool upper current of air, whose coolness has, in fiict, frozen the 

 vapor into snow. If the motion of the cloud follows the direction of the 

 niovement of these threads, or is directly opposed to it, the angle of fil- 

 ature will be zero. The actual filature angle is the resultant of the 

 movement of the upper current carrying the snow crystals in one direc- 

 tion, while the lower current carries the cloud, with its perpetually re- 

 newed summit, in another direction; by constructing the triangle or 

 parallelogram of motions, we can thus either construct the resulting, or 

 analyze the observed, filature. The observed result as above given, 

 namely, that the positive angle of filature occurs principally with the 

 movement of the clouds toward north and west, and the negative angle 

 with movement towards southeast, shows that the upper currents of air 

 are from the north and west when the lower current is from the south 

 or west, so that the axis of the cloud must then deviate from its summit 

 toward the right j on the other hand, southeast upper winds are most 



