CLOUDS. 451 



cloud of the correspondiug type. During a continuous rain, Trlien tiie 

 sky is completely covered with a palUo-cumuliis, we are sure to find above 

 it a second stratum of pallio-eirrus, which occasions this rain. So after 

 this examination we write in each quadrant unity for these two types. 

 But as soon as a breach is made in the stratum of pallio-ciimulus, then 

 care must be taken not to confound the quantity of cloud corresponding 

 to each of these two strata, which are perceived one after the other. 

 With a little attention we come perfectly to know each order of cloud 

 and the space it occupies. 



DIRECTION OF THE CLOUDS. 



"We should note in another column the direction of each type of cloud 

 corresponding to the first sixteen cardinal points of the comj)ass. For 

 this purpose, we must observe the space whence the cloud sets out, and 

 that of the opposite horizon, where it is lost. When the cloud tra- 

 verses the zenithal region, the observation can be easily made. There is 

 only a single ijosition which can give rise to error by an effect of per- 

 spective, which takes place morning and evening, when the cumuli are 

 not removed from the limits of the horizon, where they have a very slow 

 march and disappear at the opposite side in the same parallel plane. 

 AYe think then that a cloud freely sets from east to west, or vice versa, 

 either by the north or by the south, when it has rather an inclination 

 from northeast, fi'om northwest, from southeast, from southwest, or any 

 other. If it is at sunrise or sunset, if the wind is from east to west, 

 or if the vane remains stationary in one of these directions, we may be 

 certain that the onmiulus pursues this horizontal course perpendicularly 

 to the meridian. 



Often it is very difiicult to grasp the direction of cirrus, because of its 

 extreme slowness, and the considerable quantity and great extent of its 

 filaments, which are oriented on every side. The attention must be prin- 

 cipally fixed upon the side of the disiilacement of the ridge or central 

 trunk whence this multitude of bands and lateral filaments is detached. 

 The march of cirrus is then almost always in a longitudinal plane or 

 parallel to the longer axis. By a law of perspective, the parallel bands 

 appear to diverge from a point of the horizon, and on the other hand to 

 converge toward another point diametrically opposite ; but the observa- 

 tion of the point of convergence on the opposite horizon will give the 

 mode of orientation. 



There is another optical illusion, against which we must especially 

 guard, in order not to commit a very grave error ; for it appears each 

 time that below a stratum of superior and very slow cirrus, we perceive 

 a second stratum of inferior and rapid cumulus. In this case the cirrus 

 seems to march rapidly in the opposite direction to that of the cumulus, 

 when, in reality, it is following the same path, but more slowly. It is 

 an illusion analogous to that remarked in a railway carriage when the 

 objects closer to us file rapidly past in a direction contrary to that of the 



