These oloud forms are the precursors of thunderstorms, but the 

 first named variety may appear a long way in advance. 



9. Altocumulus in Several Sheets at Different Levels, Generally Asso- 

 ciated With Thick Fibrous Veils of Cloud, and a Chaotic Appear- 

 ance of the Sky (39, 40). 



This type is very difficult to analyse; it is generally complex, and 

 patches of middle cloud, more or less fragmentary, are seen super- 

 posed; they are often badly defined and with soft outlines; there are 

 all the transitional forms between low altocumulus and the fibrous 

 veil. The sky moreover is covered with clouds of difl^erent layers, 

 but as in general there is not a continuous sheet, blue patches may 

 be seen. Luckily, though these middle clouds are very difficult to 

 classify individually, the state of the whole sky is very typical ; it has 

 a disordered, heavy, and stagnant appearance (calm or light wind). 



These appearances are characteristic of the central parts of thun- 

 dery disturbances. 



The clouds of other levels which almost always exist in a thundery 

 sky are principally cirrus densus ((7^=3) and large cumulus ((7^=2) 

 or cumulonimbus {Cl=3). 



It would be too complicated, and hardly worth while, to attempt 

 to analyse in detail the individual characters of the altocumulus 

 sheets in a thundery sky. It should be carefully noted that such a 

 sky can be identified by its general appearance as a whole, which is 

 quite typical. There is only one other type which in this respect 

 may be confused with a thundery sky, because it also has a more or 

 less disordered appearance ; this is the sky in the rear of a disturbance, 

 but this latter is fairly easily distinguished by its "tossed about" 

 and "windswept" appearance. 



UPPER CLOUDS Ch 



0. No Upper Clouds. 



1. Cirrus, Delicate, Not Increasing, Scattered and Isolated Masses 



(41). 



This type of isolated cirrus is widely scattered; its amount does 

 not noticeably increase either in time or in any particular direction. 

 It does not collect into sheets or bands, and there is no tendency for 

 the elements to fuse together into masses of cirrostratus. The cirrus 

 clouds whose strands end in an upturned hook or tuft must not be 

 included in this class. 



Cirrus clouds of this type are uidications of a very distant dis- 

 turbance, and are found either on its front or on its borders. The 

 place where these clouds are observed is generally not influenced by 

 the disturbance, at any rate for some time. 



The chief characteristic of the type 0^=1 is the sparseness of the 

 cirrus. It is only distinguished from the type C„=2 by this fact of 

 the cirrus being more sparse. 



This type of cirrus is distinguished: From that of class 0^ = 3 in 

 that it is delicate and does not originate in anvil clouds; from class 

 C^=4 in that it does not increase either in time or in any particular 

 direction of the sky, and that it does not belong to the uncinus type 

 (strands with upturned ends); from class C« = 5 and class Ch=^ in 

 that it does not increase in time or in any particular direction, that 



98803" — 38 3 29 



