286 EXPERIMENTAL INVESTIGATION OF THE ACTION OF MEDICINES. 



two or three may often be taken on one paper by having two or 

 three exactly similar <]jlass pens of the form shown tilled with 

 inks of different colours. Each is stuck through a small piece 

 of cork, and into the under side of the cork a small glass tube 

 is put, which will just fit the top of the swimmer. l>y simply 

 dropping the glass tube on the end of the swimmer, the pen is 

 in its place at once, and can be changed with great facility. A 

 small sable brush may also be substituted for the glass pen. 



After the experiment has gone on for some little time, a clot 

 is apt to form in the cannula. When this is the case, the clip 

 must be replaced on the artery, the stopcock (d) turned trans- 

 versely, so as to keep the mercury at the same height, the clip 

 on the india-rubber tube at d B and at e A removed, and the 

 tube washed out by a stream of carbonate of soda. Any clot in 

 the cannula is removed by a spill of twisted paper, by a hog's 

 bristle, or by a piece of whalebone. The whalebone-probes are 

 most convenient, as they can be made of any size. A single 

 jet of blood should now be allowed to escape from the artery, so 

 as to make sure that there is no clot in it, the tube again 

 washed out with carbonate of soda, the clips at e B and e A 

 replaced, that on the artery removed, and the stopcock turned 

 and tracings taken as before. 



Bedudion of the KymogrcqjJiion Tracings. — It is not only im- 

 possible to publish the tracings as they are taken from the 

 kymographion for the benefit of others, but it is extremely 

 difficult to draw any except very general conclusions from them 

 for one's self. Before they can be of much use they must be 

 reduced to tables, or, wdiat is still better, the tables themselves 

 may be graphically represented. In making the tables, we 

 must first fix the time at which the different parts of the tracing 

 were made. The time when the tracing was begun and when the 

 injection was made must be noted down at the time in a separate 

 note-book, or, still better, on the tracing itself. In the first tracing 

 it is convenient to take the time when the injection was made, 

 as a starting-point from which to reckon the other periods. 



Beginning at this point, then, we divide the abscissa or zero 

 line into parts corresponding to five seconds each, or any other 

 period we think convenient. If the circumference of the 



