of this with light petroleum. Many resins and gum resins are used in the manufacture of varnishes in conjunction with linseed oil, but none of them can deprive the oil of the defect referred to, and if used in too large a proportion they become too brittle for balloon purposes. Both scientific investigation and practical experience show that any varnish containing linseed oil must be looked upon with suspicion by the aëronaut, in spite of the glowing testimonials some manufacturers are always ready to give their own goods. When we consider those varnishes which are solutions and which do not depend upon oxidation for their drying properties we enter upon a very wide field. Practically any substance that is soluble in a neutral solvent and leaves an impermeable film on drying is included in this class. One of the simplest examples is gelatine in its various forms, with water as a solvent. Until recently glue or gelatine would have been useless for our purpose on account of its ready solubility in water, but now that we are able to render it insoluble by means of chromic acid or formaldehyde it comes within the limits of practical applicability. A fabric may be rendered almost impermeable to gas when coated on the inside with insoluble gelatine, and on the outside with a waterproof varnish. Animal membranes are far less permeable to gases than fabrics coated with varnishes of the usual kinds. A balloon of gold-beaters’ skin, if carefully constructed, will retain hydrogen gas for a long time, and if treated with gelatine that is afterwards rendered insoluble it becomes practically impermeable. Fabrics treated with linseed oil varnish, on the other hand, allow gas to pass with comparative ease. This is not a question of porosity or “pinholes,” as is sometimes imagined, but a property inherent to the material. Hydrogen or coal gas is absorbed on the one side of the film and given off on the other in the same way as carbonic oxide will pass through cast iron. An inert gas, such as nitrogen, does not appear to diffuse in this way, even when there is a considerable difference in pressure between the two sides of the film. Such a varnished fabric transmits hydrogen readily, but retains nitrogen, and is perfectly watertight. In filling up the interstices of a fabric composed of cellulose the most obvious substance to use would be cellulose itself, but until recently solutions of this kind were difficult to obtain. Toy balloons have long been made of collodion, and are fairly satisfactory, but a cotton fabric impregnated with pure collodion becomes hard and even brittle. Celluloid solution, which is collodion with camphor and a small quantity of castor oil, is more flexible, but, probably on account of the camphor, is more permeable to hydrogen than collodion. A