HISTORY OF CHRONOPHOTOOKAI'HY. 339 



In the construction of my new apparatu.s the section of the air tube 

 was increased from 20 to 50 cm. and the number of threads of smoke 

 from 20 to 5S. The filtering cloths were replaced ])y silk gauzes with 

 a very small mesh, and I finally introduced into the experiment a 

 chronographic system which allows us to measure the speed of each 

 smoke thread in diti'erent parts of its course. For this purpose the 

 system of little tubes which liring the smoke threads which are about 

 to be aspii'ed is subject to a lateral shake, repeated ten times every 

 second. An electric vibrator regulates this movement with the above- 

 named frequency, and under this intluence the smoke threads do not 

 form straight, parallel lines, but sinusoidal curves. These inflections 

 are preserved during their whole path. In the interior of the tube a 

 small scale 20 cm. long, in the same plane as the smoke threads, serves 

 to measure the space traversed by the molecules of air in each tenth 

 of a second. 



Some examples of the results obtained will enable us to appreciate 

 the progress which has ])een made in the new construction. 



When there is no obstacle ofi'ered to the air current the smoke 

 threads remain rectilinear and parallel. If we place an inclined plane 

 in the current the smoke threads enlarge in meeting it, which indicates 

 that they lose velocity before following opposite directions. Some 

 mount toward the upper edge of the plane, others glide upon each 

 other without mingling and escape by the lower edge. On each side 

 of the obstacle the smoke threads continue their motion very close 

 together, leaving behind the inclined plane a space where the air is 

 motionless, and only gives a smoky cloud. This space where the eddies 

 or whirlpools occur is larger in proportion as the obstacle to the air 

 current is larger. 



To note the speed of the air current in diflerent parts of its course 

 we repeat the experiment, sul)jecting the smoke threads to the above- 

 mentioned vibrations, and then the threads instead of being rectilinear 

 present a series of lateral inflections which are preserved during all 

 their course. These inflections remain equi distant if the speed of the 

 current is everywhere the same, but if the current speed diminishes 

 the inflections are closer; if it is rapid, the}^ are more distant from 

 each other, and the space moved over in a given time is measured by 

 means of the metric scale. 



The figures which we have just seen are observed t)y a magnesium 

 flash; that is to say, in so short a time, that each smoke thread seems 

 immovable. If the light lasted longer the aspect of the figure would 

 change and give the further condition of the air current as we see it 

 in fig. 4'', where the light produced by the prolonged combustion lasts 

 about seven seconds. 



"TJiis rt^. 4 corresponds to fig. 2i), PI. v, of the foregoing paper on i-lironophotog- 

 raphy, where are also shown other figures here referred to by Doctor Marey. 



