Jo UNIVERSITY OF MISSOURI STUDIES 



the length of the partition. The partition as a whole is much 

 narrower in comparison to its length than appears in the fig- 

 ure. Let us try, then, to get a clear conception of the func- 

 tional significance of these facts. It is of no particular im- 

 portance, in this connection, whether the measurements upon 



Fig. 24. Shape of the partition 



which the following considerations are based are more or less 

 incorrect, as they probably are ; for our intention is merely to 

 get an idea of the general direction in which the actual shape 

 of the partition changes the results of a theory having pro- 

 visionally assumed that the partition is everywhere of equal 

 width. 



When the partition yields in either direction, up or down, 

 its former place is taken by the fluid of the tube. Let us call 



the quantity of fluid which has taken po- 

 A unit of stirruD sitions formerly occupied by the partition 

 movement equals "the displaced fluid." Now, it is plain that 

 a unit of dis- the quantity of displaced fluid must al- 



placed fluid ways be approximately proportional to the 



distance through which the stirrup has 

 moved since its last reversal of movement. If the partition 

 were equally wide ever3^where, then any section of equal 

 length, far from or near the windowis, would make room, in 

 moving from one limit to the other, to the same quantity of 

 displaced fluid as any other section. And then, plainly, the 

 length of that part of the partition which is caused to move 

 from one limit to the other would always be proportional to 

 that part of the stirrup movement which caused it to move. 



