Mr. W.S. Jevons on the Cirrous form of Cloud. 31 
of the glass cylinder passing before it; and being a good con- 
ducting body, the cloud soon discharges itself to the earth in a 
succession of immense sparks or flashes, which will be kept up 
as long as a current of moist, warm and excited air continues to 
arrive at the storm-point, or this latter itself travels over new 
and excited regions of air. The continual supply of electricity, 
often amazingly great, at which some authors have much won- 
dered (see Miller’s ‘ Elements,’ vol. i. p. 318), would at all events 
be easy to understand ; and this explanation can be extended to 
the phenomenon of lightning in any cumulous cloud, since the 
cirrous crest or the stratous masses which also usually form, are 
in no way essential to the collection of electricity. 
The section of a thunder-cloud, on the preceding page (fig. 3), 
is only intended to represent a certain class of storms occurring in 
Sydney, since other storms of different character occur even here, 
but it seems to agree exactly with the descriptions of storm- 
clouds in other localities (see Arago’s ‘Essay on Thunder and 
Lightning,’ Chap. II.; or Howard’s ‘ Climate of London,’ Intro- 
duction, p. xlviii). Arago proves that lightning may issue from 
a single small cloud; but in this case I presume it will always 
be a cumulous cloud, both from the descriptions of the clouds 
in the cases cited, and from the remark of Beccaria, which he 
quotes, viz. that “thunder and lightning never issue from smoky 
clouds; that is to say, from those strata of clouds which are 
characterized by their apparent uniformity of composition and 
regularity of form,” meaning, I suppose, stratous, cirrostratous 
or cirrous clouds. 
It will perhaps have been observed in our experiment No. 1, 
that the streams descending from the upper stratum into the 
lower often end in little knobs, or drops, or scrolls of a peculiar 
and interesting shape. I do not understand why the descending 
streams should differ in shape from the ascending ones, which 
generally, but not always, end in evanescent points, though it 
might arise from the tendency of chloride of silver to subside, as 
mentioned before; but it is remarkable that similar appearances 
are often to be seen on the under surface of dense cirrostratous 
clouds, especially at the front or the tail of a thunder-cloud (as 
shown in figure at E). Sometimes these dropping portions of 
cloud, or droplets, seem to come in contact with dry air, when 
their well-defined form is destroyed, and a fibrous or fur-like 
appearance only remains. They appear to be truly portions of 
subsiding cloud. 
To return to our theory, Howard says (‘Climate of London,’ 
Introduction, xlii), ‘‘ Steady high winds are also preceded and 
attended by streaks (of cirrus) running quite across the sky in 
the direction in which they blow.” This would be the precise 
