1878.] 
Correspondence. 
23, 1871), that optical phenomena of 
the kind here referred to indicate a 
transition state of the weather ; and 
this inference has recently been again 
substantiated (Odtober, 1877) by the 
occurrence, on two successive days, 
of imperfedt parhelia at the end of a 
period of fine weather, and before the 
advent of storms and rain. That 
this is the true prognostic value of 
these appearances will scarcely be 
disputed now by meteorologists. How 
to account for the fadl that ice-crystals 
become chiefly visible at these times 
of transition does not appear easy. 
It is well known that in the finest 
weather abundance of crystals may 
be found at high elevations, even with 
a clear sky ; but it seems perfectly 
clear that optical phenomena do not 
always attend them. 
I venture, therefore, to propose the 
following explanation of the fadl that 
halos generally attend a “ transition ” 
state of the atmospheric vapour, viz., 
that they result from the transmission 
141 
of the solar rays through their sub- 
stance while in a partially melted 
state, or, rather, while slightly melted 
on the surface. 
It must be observed that in dry 
frosty weather the longer ice remains 
exposed to atmospheric influences 
the more uneven its surface becomes, 
by the accretion of vapour upon its 
prominences, and consequently the 
less capable it becomes of giving a 
clear refradtion and exhibiting chro- 
matic effedts. Now if we allow that, 
previous to the appearance of a halo 
or parhelion, the floating ice may have 
been present in this condition, a rise 
of temperature would produce the 
optical phenomena referred to, by 
making the refradtion more perfedt ; 
and, on the other hand, a fall of tem- 
perature sufficient to freeze would 
have the same effedt, by origination 
of new ice-prisms from existent 
watery vapour. — I am, &c., 
S. Barber. 
Tilmanstone, Sandwich. 
