364- Formation and Characteristics 



of cirro-cumulus are approximated in groups, while those of 

 cumulus are isolated. Indeed, when the masses of cumuli 

 just mentioned have a tendency to approach each other, they 

 gradually form the variety of cirro-cumulus above noticed ; 

 and I have observed several of the larger masses of this 

 modification of cumulus separate into the larger masses of 

 cirro-cumulus ; the connexion, therefore, between these modi- 

 fications may be thus expressed : cirrus, cirro-stratus, cirro- 

 cumulus, cumulus. 



In many instances, the cirro-stratus is formed from cirrus : 

 not by a conversion of the cirrus into this kind of cloud, but 

 merely by an augmentation of the fibres composing cirrus. 

 It is also produced immediately from cirro-cumulus; the small 

 roundish masses of the latter cloud changing into the fibres 

 that characterise cirro-stratus : this process being the reverse 

 of the resolution of cirro-stratus into cirro-cumulus, which 

 is frequently observed to take place. In the latter case, the 

 cirro-cumulus is produced by a conversion of the fibres of 

 the cirro-stratus into the well-defined masses of cirro-cumulus. 

 This conversion has been observed at all altitudes, and when 

 the clouds have exhibited every possible degree of size : the 

 cirro-stratus previously possessing a fibrous structure with a 

 longitudinal arrangement, these conversions have taken place, 

 in many instances, very rapidly. 



The connexion between cirro-cumulus and cumulus has 

 been explained above. 



From the whole of my observations, it appears to me, that 

 cirro-stratus and cirro-cumulus are the only modifications of 

 cloud that interchange (and this phenomenon is confined to 

 certain varieties) ; that cirro-stratus is produced, in some in- 

 stances, by an augmentation of the fibres composing cirrus ; 

 and cirro-cumulus, on some occasions, by an approximation to 

 each other, and an increase in bulk of the masses of a certain 

 variety of cumulus. I do not remember having observed 

 cirrus result from cirro-stratus, nor cumulus from cirro- 

 cumulus. On some occasions, I have seen certain varieties 

 of cirro-stratus skirted with cirrus ; but I have not seen the 

 former cloud resolve into cirrus, as it often does into cirro- 

 cumulus. 



The polarisation of clouds forms an interesting feature in 

 their natural history : by polarisation I mean the direction, or 

 stretching from one point of the horizon to the opposite (s.w. 

 to n.e., for instance), of the groups of cirrus, cirro-stratus, 

 and cirro-cumulus, which are the modifications most com- 

 monly polarised. Their polarisation occurs in all directions, 

 and appears to be entirely unconnected with their motion, 



