KNOWLEDGE 



[Oct. 13, 1882 



CLOUDS IN THE AIR. 



By the Editor. 



(Ci>n«i»u<d /rom page 279.) 



IN our last wp considered the effect of foreshortening on 

 roanded clouds ; we have now to consider how those 

 clouds which arc really long, parallel bands appear, as seen 

 from different points of view. As before, wo imagine clouds 

 and arrangements of clouds altogetlier more regular than 

 in nature. In fact, it is only thus tliat the problems we 

 Lave to deal with can be effectually treated ; for if we gave 

 to the clouds we considered peculiarities of shape such as 

 actually e.xist in nature, it would be impossible to dis- 

 tinguish properly the effects of such peculiarities from the 

 -effivts due to the position of the clouds, or cloud-bands, 

 ■%k-ith respect to the observer. 



Let us suppose, in the first place, that the clouds are 

 arranged in a series of parallel bands, as shown in Fig. 1. 

 Tliese clouds may be supposed to be roughly circular in 

 section athwart their length. The portion of sky supposed 

 to lie shown in the figure may be regarded as about the 



having, in fact, the same form as the outline of one of the 

 bands shown in Fig. 'J, about midway been the centre and 

 the edge. Wo see from Fig. 1 that tliis does not really 

 show, as it seems to do, that the cloud is higher in the 

 middle than at eitlit-r side, but that tlio part of the edge 

 which seems highest is nearer than the part which looks 

 lower. Thus when a part of the under edge of a cloud-bank 

 which we are npproiiching soenis liigher than the rest, we 

 may iuf.r tliut tliiit jiart is neanr than the rest, and that 



same as is shown in the three last figures of the last article 

 (pp. 278 and 279), or alnjut forty-five degrees from the 



-zenitli, all round. [But lat<r, when we have to consider the 

 effect* produced by for<!8hortening near the horizon, we 



tnay aliio regard the picture as showing the whole atmo- 

 sphere alx)ve a place.] 



Li Fig. 2 tlie cloud-ljands, shown in Fig. 1 as they 

 actually are, are shown as they appear to an eye placed 

 immediately Vx:neath the region reprejiented. In other- 

 words, the centre of the figure represents the point over 

 head, while the circumference represents a circle on the 

 sky al>out forty-five degrees above the horizon. We see 

 that the bands, which are really straight, app<;ar to the eye 

 to l>e curved. It may l<c noticed here that this figure 

 '■n&bles UR also to recognise the real meaning of the curved 

 outline ofUrn seen in the case of a cloud covering the 

 region over head. Such a cloud will oft<;n be seen to have 

 an outline like an arch, drawing close to the horizon on 

 either side, and rising high alxive the horizon in the middle — - 



Fig. 2. 



the sky overhead will soon (if wo continue our advance) 

 be free from clouds. 



But while the parallel bands of clouds in the region 

 overhead an? thus modified in form, a much more marked 

 change atiects parallel bands lying near the horizon. Of 

 course, the nature of the change depends in part on the 

 direction of the bands. If we look towards that part of 

 the horizon where the length of the cloud-baiids is at right 

 angles with the line of sight, we see the bands all brought 

 together by foreshortening, so as to form a mass of stratus 



cloud, marked by a series of striations parull(!l to the 

 horizon. If, on the other hand, we look towards the 

 liorizon in a direction parallel to the cloud bands, a very 

 different efl'ect is produced. For then, by a well known 

 •effect of perspective, the vanishing point of the parallel 

 bands is on the horizon, and the clouds appcsar to radiate 

 from the horizon, as shown in Fig. .3. It is obvious that 

 the same effect will l^e produced whether w(! look in cither 

 of the two directions parallel to the cloud-bands — so that, 

 when the whole of the visible sky is occupied by parallel 



