i8o 



STUDIES IN LUMINESCENCE. 



deviations from the expected proportionality between the slit-width and 

 intensity, especially for widths of less than o.oi cm. (see Figs. 163 and 164 

 in Chapter XI). 



It was therefore decided to avoid changes in slit width by using the 

 method of comparison used in determining the distribution of energy in the 

 spectrum of the comparison flame. In this method the two slits of the spec- 

 trophotometer remain unchanged in width, and equality of brightness is 

 obtained for each region of the spectrum by moving the comparison flame 

 along a bar or track parallel to the axis of the collimator. With the system 

 of screens which we employed to exclude stray light, the law of inverse 



squares was found to hold 

 for the entire range of dis- 

 tances used in our experi- 

 ments. 



The three substances 

 selected for measurements 

 were fluorescein in aque- 

 ous solution slightly alka- 

 line, eosin in alcohol, and 

 resorufin in alcohol. 



The solutions in each case 



>Hl 



* * 



I 



6 H - 



Fig. 175. 



were as dilute as was found practicable, so as to reduce the correction for 

 absorption to a minimum. 



In the determination of the fluorescence curves the solution was placed 

 in a rectangular cell of white glass (/, Fig. 175) and was excited by the light 

 from a Cooper-Hewitt mercury lamp, C. II. The tube of this lamp was 

 vertical and mounted at a distance of about 30 cm. from the wall of the cell. 

 Only those portions of the tube were used which were nearly in the same 

 horizontal plane as the cell. The beam of exciting light entered the cell/, 

 Fig. 175, in the direction of the arrow at right angles to the axis of the 

 collimator. 



The cell was inclosed within a metal box with black, matte, oxidized 

 surfaces, and having only the broad rectangular opening de for the admission 

 of the exciting light and a narrow, vertical, slit-like aperture opposite the 

 slit b through which the fluorescence was viewed. 



The use of the mercury arc, with its almost complete absence of light in 

 the region occupied by the fluorescence bands to be measured, afforded 

 further protection against stray light. In the study of the fluorescein solu- 

 tion the additional precaution was taken of inserting a cell of ammonio- 

 sulphate of copper in water between the lamp and the fluorescent liquid, 

 thus cutting off the yellow and green lines of the arc almost completely. 

 The mercury lamp was fed from a storage battery of 120 volts with suitable 

 resistance in series, and under these conditions it furnished an exciting light 

 of unexpected constancy, surpassing in this respect any other source of suit- 

 able character and sufficient intensity with which we have had experience. 



The arrangement of the apparatus for determining the fluorescence spectra 

 is shown in Fig. 175. The plan is similar to that used in comparing the 

 acetylene flame with the black body; but the photometer track carrying 

 the flame was mounted in line with collimator a, while the fluorescence cell 



