The Extant Letters 23
position and were unable to consider together the phenomena which one
must accept as signs.
- The varying lengths of nights and days [could occur] as a result
of the alternate swift and slow motions of the sun over the earth, <or
even> as a result of covering the varying distances between places and
certain places either faster or slower, as is also observed [to happen] with
some things in our experience; and we must speak in a manner consistent
with these when we speak of meteorological phenomena. But those who
accept one explanation are in conflict with the phenomena and have lost
track of what it is possible for a man to understand.
Predictive weather signs could occur as a result of coincidental conjunc-
tions of events, as in the case of animals which are evident in our experi-
ence, and also as a result of alterations and changes in the air. For both
of these are not in conflict with the phenomena; 99. but it is not possible
to see in what sort of cases the explanation is given by reference to this
or that cause.
Clouds could come to be and to be formed both as a result of thickenings
of air caused by the pressure of the winds, and as a result of the entangle-
ments of atoms which grip one another and are suitable for producing
this effect, and as a result of a collection of effluences from both earth
and bodies of water; and it is not impossible that the formation of such
compounds is also produced in several other ways. So rains [lit. waters]
could be produced from the clouds, sometimes when they are compressed
and sometimes when they undergo change; 100. and again, winds, by
their egress from suitable places and motion through the air, [can cause
rain] when there is a relatively forceful influx from certain aggregates
which are suitable for such discharges.
Thunder can occur as a result of the confinement of wind in the
hollows of the clouds, as happens in closed vessels [in] our [experience],
and as a result of the booming of fire combined with wind inside the
clouds, and as a result of the rupture and separation of clouds, and by
the friction between clouds and their fragmentation when they have taken
on an ice-like solidity. And the phenomena invite us to say that the entire
topic as well as this part of it are subject to several different explanations. - And lightning flashes similarly occur in several different ways;
for the [atomic] configuration which produces fire is squeezed out by the
friction and collision of clouds and so generates a lightning flash; [it
could] also [occur] as a result of the wind making the sort of bodies
which cause this luminiscence flash forth from the clouds; and by the
squeezing of clouds when they are compressed, either by each other or
by the winds; and by the inclusion [in them] of the light scattered from