Theory of the Pneumatic Paradox. 311 



A disk of card two inches in diameter, is placed horizontally at 

 the upper end of the tube. Another card of the same size made 

 slightly concave, is placed over the first. Let the whole be cov- 

 ered with a glass, receiver ; then let the air be removed, which of 

 course can only quit the receiver by passing out through the brass 

 tube ; when the exhaustion is carried to the utmost limit, let the 

 air be readmitted in a full gush through the brass tube, and the 

 upper card will not be blown off. This experiment has been 

 tried many times with various states of exhaustion, from a quarter 

 of an inch of mercury to twenty inches, but in no case was the 

 card blown off, or even agitated in the slightest degree ; therefore 

 the upper card is not retained in its position by the pressure of 

 the atmosphere." 



Mr. Tomlinson's failure to blow off the disk may have arisen 

 from the concavity of the upper card, from not carrying the ex- 

 haustion far enough in proportion to the size of the movable disk, 

 or from the warping of the cards, which, owing to the evapora- 

 tion of moisture usually contained in their substance even when 

 apparently quite dry, is very apt to take place as the exhaustion 

 proceeds. To obviate this difficulty, the fixed disk may be made 

 of tinned sheet iron or brass, and the movable one of the same 

 material, very thin and light, or of card recently dried by expo- 

 sure to heat, and pressed, in order that both disks may be perfectly 

 plane and in contact throughout their whole extent, when applied 

 to each other. For the convenience of suspending, during the 

 process of exhaustion, the movable disk from the top of the re- 

 ceiver, by means of a rod sliding through a collar of leathers, and 

 also to prevent it from moving laterally out of place, a cambric 

 needle may be passed through its centre, projecting, when the two 

 disks are applied to each other, about an eighth of an inch into the 

 • tube. I repeated Mr. Tomlinson's experiment with a very pow- 

 erful air-pump, recently made for the University of Cambridge 

 by N. B. Chamberlain of this city. The passage by which air 

 is admitted into the receiver is one eighth of an inch in diameter. 

 Every thing having been adjusted as above described, the air was 

 exhausted to such a degree as to reduce the mercury in a very 

 accurate syphon gauge to within one tenth of an inch of a level 

 in the two branches, indicating a residue in the receiver of only 

 one three hundredth of the air naturally contained in it. Bj 

 means of two stop-cocks the air was first admitted into the air pas- 



