now enlarge the lower chamber so that the expansion may be in ratio of 500 to 760 or 1 to 1.5. Moreover, I will not allow any dust to enter the upper globe, but will draw into that globe dustless air filtered through this bit of cotton wool. If now I allow moist or saturated dustless air to expand into the lower chamber from the pressure 760 to that of 500 or an expansion of about 1.5 in volume then I shall form not a cloud of small particles, but a few larger drops of water. This is the process that must be going on within the thunder-cloud, or in fact inside any raincloud. Out of the great mass of moist air that makes up the whole cloud only a small proportion is free from dust and of that only a small portion expands rapidly enough to form drops of rain-water.
I think you will see that the firing of cannon or dynamite in order to make a great noise is not likely to form rain and in fact can not possibly bring it down. Neither can it prevent the formation of hail or rain. If we wish to avert heavy rain or hail we must either cut off the supply of moisture, or else prevent the rapid expansion, or else throw dust upward into the air to cause cloudy condensation instead of rain. Apparently this latter process is carried out for us in nature when great forest fires afford enough particles of smoke to provide for the cloudy condensation of the free moisture. From the great clouds of smoke that attend these forest fires we get no rain until after a long time the heat that is in the cloud is lost by radiation, or until larger drops are formed by further expansion.
Even the bombardment of a cloud by the explosion of dynamite within it is inefficient to produce rain, in part because no violent concussion can drive the cloud-particles together into large drops of rain, and in part because the explosive itself furnishes more dust particles and more nuclei of condensation and therefore produces clouds instead of rain. In the same way the bombardment of the clouds by means of vortex rings is inefficient.
I have here an apparatus for making vortex rings of air; you notice that a slight stroke on the rear side of this box drives forward a vortex ring of smoky air. It is a beautiful sight and very instructive in many ways, but the special form of cannon devised in Italy to send such rings of gunpowder smoke up into the clouds and break up the formation of hail does not usually send them higher than 1,000 feet; they break up long before they reach the clouds. We have no evidence that they ever reach them, or that they could have any effect if they did so. If they carry up much dust they ought to have a slight effect in producing cloudy condensation, and thus cutting off the formation of rain or hail. But this effect is certainly too slight to be appreciable in our statistics. I regret to think of so many thousands of farmers wasting time and money on this delusion. You know that De Morgan after spending much time in combating analogous delusions wrote an inter-
