WILLIAM R. AMBERSON 521 



effort has been made to control the whole manipulation of this film 

 so as to secure the greatest possible uniformity. To facilitate this 

 end only short strips of film have been used, in which great vari- 

 ations in the emulsion would not be expected. The standard length 

 of record finally adopted is 205 mm. long. The whole length of 

 film used, including the records of the reaction, and the calibration 

 exposures, is 250 mm. 



After several unsuccessful attempts to construct some optical 

 system which would function to focus the light given out by the 

 reaction upon a moving film, I turned to the use of a simple test- 

 tube container to the side of which is cemented a phosphor-bronze 

 strip in which a vertical slit window has been cut. The two con- 

 tainers used throughout the work were carefully chosen to be of 

 exactly the same bore. The two slit windows were cut in the phosphor- 

 bronze strips by a milling machine, to a width of 1.5 mm. and a 

 height of 6 mm. In width they differ by an error of considerably 

 less than 1 per cent. The two bronze plates are cemented to the 

 containers in identical positions, and were so applied that absolutely 

 clear glass only is apposed to the slit openings. The inner edges of 

 each slit come flush with the glass of the container. All outer surfaces 

 of both tube and brass plate are painted a flat black to prevent 

 transmission or reflection of light. Light can pass from the containers 

 only through the sHt windows themselves. The inner edges only 

 of these openings are unpainted, since I wished to avoid the intro- 

 duction of large errors in sHt widths which might arise in the appli- 

 cation of such coverings. 



With this simple arrangement it is possible to follow the decay of 

 light intensity in reacting Cypridina solutions with considerable 

 accuracy. The two containers, mounted upon horizontal supporting 

 bars, adjustable in length, are clamped rigidly upon uprights, and 

 are brought close to the drum of a Zimmermann kymograph. They 

 are adjusted for each experiment so that the line connecting container 

 center and drum center passes through the middle of the slit window, 

 giving a symmetrical and identical orientation for both containers 

 wdth respect to the drum. The two containers are placed side by 

 side, one of them being fixed about 10 mm. higher than the other. 

 The slit windows are thus presented to two bands of the drum surface 



