August 2, 1895.] 



SCIENCE. 



123 



blown off every time. The size of A B, 

 Figs. 2 and 3, varies as the square of the 

 radins of G, and as the distance of A B 

 from C D. 



This seems a sufficient confirmation of the 

 theory of the action of the first instant of 

 time. 



The subsequent action can be explained 

 as follows : Fig. 5. The current from G, 

 rising strikes A B on E F and spreads out 

 in a direction M, passing on along M at 

 some point H, the current M forms a whirl- 

 pool H and spreads, taking most or all of 

 the air beyond H slowly along with it, out 

 from between A B and C D. The currents 

 G and M gradually attract and draw with 

 themselves a part of the dead air in E and 

 form a current K. The tendency of a cur- 

 rent is to draw into, and along with itself, 

 the adjoining dead fluid through which it 

 flows. M and H thus drawing the air from 

 R start a current L, which gradually draws 

 out air from R, thus caiising a slight vac- 

 uum directly over O N, and the pressure of 

 the atmosphere on O N tends to press the 

 two cards together to fill the space H. C 

 D cannot rise, being fast to the tube, so A 

 B is bent down iintil the space between F 

 and O is just sufficient for the air from G to 

 pass out through, i. e., when A B and C D 

 are separated by a distance equal to one- 

 half the radius of G. This is shown in Fig. 

 4. The instant this limit is reached the ac- 

 tion of Fig. 5 is replaced by that of Fig. 4, 

 and the card by its own elasticity and the 

 pressure over G rises and the action of Fig. 

 5 is restored. Thus in almost all cases the 

 card vibrates rapidly. The experiments to 

 confirm this theory were made almost en- 

 tirely with water, since the results and ac- 

 tions would be the same and it was more 

 convenient. The apparatus used consisted 

 of a piece of thin board with four upright 

 wires near the corners, upon which slid a 

 second board; this was so arranged in order 

 that the distance from A B to C D could be 



varied at pleasure. Instead of introducing 

 the tube G through the middle of the board 

 C D, it was placed in the middle of one of 

 its sides, and over this edge was placed a 

 plate of glass. This arrangement was to 

 obtain a view of the currents in a section 

 through the center of G and perpendicular 

 to A B and C D. Placing this apparatus 

 in a tub of water, with the plate of glass 

 parallel and near to the siirface, and pouring 

 muddy or colored water through G, the 

 currents take the directions represented by 

 the arrows in Fig. 5. To further test the 

 action of the currents M and H in drawing 

 the water from R, a bubble of air was intro- 

 duced in front of G before the current 

 started, and on starting the current in G the 

 bubble separated and each half assumed the 

 shape shown in Fig. 6, E R, and the air 

 was rapidly drawn out in little bubbles at 

 L, and driven out in a direction I. To con- 

 firm the theory that the force which holds 

 the cards together is applied at O N of the 

 bottom card, and in order to find where the 

 currents increased and where they dimin- 

 ished the pressure of the atmosphere, a sur- 

 face of water was substituted for the lower ■ 

 card, and, again obtaining a sectional view 

 with the plate of glass, it was found that 

 the surface of the water assumed a shape 

 whose section is shown in Fig. 7 ; thus 

 proving very conclusively that the pressure 

 of the atmosphere upon the surface of the 

 water was diminished to some little extent 

 by the currents M and H, and that here we 

 have the whole cause of the apparently 

 strange action of the card. On substituting 

 a surface of water for the upper card A B, 

 a section shown in Fig. 8 was obtained; thus 

 showing that the only effect of the currents 

 upon A B is the pressure at E F. These 

 four experiments. Fig. 5, with colored water 

 and Figs. 6, 7 and 8, seem to sufficiently es- 

 tablish the above theory of the cause of the 

 card being drawn down after the first in- 

 stant. The theories of the action at the be- 



