CONSTRUCTION OF BALLOONS. 83 



the too near approximation of the basket to the body of the para- 

 chute ; and especially in the want of a perpendicular cord passing 

 from the car to the centre of the concave of the umbrella, by the 

 absence of which the velori'y of the descent is certain to be very 

 rapid before the machine becomes at all distended ; whereas, if a 

 cord were thus disposed, the centre of the parachute would be the 

 portion first drawn downwards by the appended weight, and the 

 machine would he almost immediately at its full extension. Having 

 found, by experiment, the diameter, required for insuring safety, 

 the further the basket or car is from the umbrella, the less fear shall 

 we have of an inversion of the whole from violent oscillations ; yet, 

 the longer the space between the car and the head of t!ie machine, 

 the longer will be the space run through in each vibration when 

 once begun, yet by so much the more will they be steadier ; and 

 this ought to be attended to, as when by the violence of the oscilla- 

 tions the car became (in Garnerin'* experiment) on a line with the 

 horizontal axis of the machine (or, in other words, the point of sus- 

 pensatiou,) the force of gravity, or the gravitating power of the 

 weight in the car, on the umbrella, being at that crisis reduced to 

 nothing, the slightest cause might have carried the body of the 

 machine in a lateral direction, reversing the concavity of the um- 

 brella, and M. Garncrin, perhaps, have fallen upon the now convex 

 yet internal portion of the bag, and the whole have descended con- 

 fusedly together. It now remains to give some account of the 

 method by which aerostatic machines may be tilled ; and here we 

 are able to determine with much greater precision concerning the 

 inflammable-air balloons than the other kind. With regard to these, 

 a primary consideration is, the most effectual and cheap method of 

 procuring the inflammable-air. It will be found that the most ad- 

 vantageous methods are, by applying acids to certain metals ; by 

 exposing animal, vege able, and some mineral substances, in a close 

 vessel to a strong fire j or lastly, by transmitting the vapour of cer- 

 tain fluids through red-hot tubes. For obtaining inflammable-air 

 from pit. coal, asphaltum, amber, &c. &c. Mr. Cavallo recom- 

 mends the following apparatus : let a vessel be made of clay, or 

 rather of iron, in the shape of a Florence flask, somewhat larger, 

 and whose neck is longer and larger. Put the substance to be 

 used into this vessel, so as to fill about four-fifths or less of its 

 cavity. If the substance be of such a nature as to swell much 

 by (lie action of the fire, lute a tube of brass, or first a brass and 



