NO. 2 METHOD OF REACHING EXTREME ALTITUDES 25 



held apart by iron wires bound into the fencing. This construction 

 allowed most of the gas to penetrate the spiral to a considerable 

 distance before being disturbed, and, of course, eliminated regular 

 reflection. The second device, J', placed under the first, consisted 

 of a number of 12-inch circular disks of the same fencing, bound to 

 two quarter-inch iron rods, O, by iron wires. These disks were 

 spaced one inch apart. The three upper disks were single disks, the 

 next lower two were double, with the strands extending in different 

 directions, the next two were triple, and the lowest disk of all, two 

 inches from the bottom of the tank, was composed of six individual 

 disks. This lower device necessarily offered large resistance to the 

 passage of the gas ; yet strong rebound from any part of it was pre- 

 vented by the spiral just described. With this third arrangement, 

 small rebound was to be expected, which also was borne out in 

 practice. 



This tank was exhausted by way of a stopcock at its lo'wer end, S ; 

 and air was also admitted through this same stopcock. 



The circular tank, plate 8, figure 2, was 10 feet high and weighed 

 about 200 lbs. It consisted of a straight length of 3-inch pipe, care- 

 fully fitted, and welded autogenously, to a four-foot, 3-inch, U-pipe. 

 The straight pipe entered the U-pipe on the inner side of the latter, 

 and at as sharp an angle as possible. Another similar U-pipe, was 

 bolted to the first by flanges, with y^g-inch sheet rubber packing 

 between. 



In this tank, the gases were shot down the straight pipe, entered 

 the upper U-pipe at a small angle, thus avoiding any considerable 

 rebound, and thence passed around the circular part — not returning 

 up the straight pipe until the velocity had been greatly reduced by 

 friction. 



In order to make the time, during which the velocity was being 

 reduced, as long as possible, the pipes were carefully cleaned of 

 scale. They were first pickled, and then cleaned by drawing through 

 them, a number of times : first, a scraper of sheet iron ; second, a stiff 

 cylindrical bristle brush, and finally a cloth. All but the most firmly 

 adhering scale was thereby removed. Further, care was taken to cut 

 the hole in the rubber washers, between the flanges, so wide that 

 compression by the flanges would not spread the rubber into the pipe, 

 and thereby obstruct the flow of gas. 



Notwithstanding all these precautions, evidence was had that the 

 gases became stopped very rapidly. This was to be expected inas- 

 much as there is solid matter, namely, the wadding and wire, that is 



