NINETEENTH ANNUAL MEETING. 



41 



coloring matter used in this specimen is probably indigo, but the plaster of paris 

 and magnesium carbonate were undoubtedly colored with ultramarine, in conse- 

 quence of which the entire violet end of the spectrum is much brighter than in 

 a true white. Thus the spectrophotometer affords a means of showing the presence 

 of such coloring matters in quantities much too minute to be detected in any other 

 way. 



The preponderance of the longer wave-lengths in the spectrum of unadulterated 

 white pigments is not due to selective reflexion, but is produced by absorption of the 

 more refrangible rays in process of transmission, so that those portions of the light 

 which penetrate to a considerable depth and are reflected internally, acquire a very de- 

 cidedly reddish tone, which although diluted by mixture with the unmodified light re- 

 flected at the surface, gives the creamy tint common to such substances. That this 

 is true can often be shown by coating a piece of glass with the substance in question 

 in the form of powder. Such a coating, when of considerable density, is found to 

 transmit light of a strong orange or red color, which disappears whenever the coat- 

 ing becomes sufficiently thin. The appearance of such coatings by reflected light 

 shows the modification due to the admixture of the red rays reflected internally, 

 sometimes in a very striking manner, and the color, as the coating becomes thinner 

 and thinner, invariably tends toward blue. In the case of some substances, notably 

 of the well-known 

 blowpipe films of 

 antimony oxide, the 

 blue appearance is 

 very pronounced 

 The spectrophoto- 

 metric analysis of 

 these blue films and 

 of the blue precip- 

 itate in very dilute 

 opalescent solu- 

 tions — for instance, 

 the solution of hy- 

 persulphate of soda 

 from which the sul- 

 phur has been pre- 

 cipitated by an acid 

 — led to a surprising 

 result. The fumes 

 of antimony before 



the blowpipe, having been caught upon the surface of a pane of white glass, showed 

 when viewed by reflected light the usual appearance of a white film shaded off by in- 

 sensible gradations, as the coating became thinner, to a delicate sky-blue. By trans- 

 mitted light the thicker portions of this film were orange, the color disappearing 

 until in those parts which presented a blue appearance by reflected light, no absorp- 

 tion effect could bs detected. 



The absorption spectrum of the denser part of this film and the reflexion spectra 

 of the thickest and thinnest portions were subjected to a photometric analysis. 



The result is shown in Fig. III. 



The thin film, in spite of its unquestionably blue appearance to the unaided eye, 

 gave a very faint spectrum in which no excess of blue or violet rays could be de- 

 tected with certainty, the energy curve approaching more nearly that of a true 

 white than did the curve of any so-called white pigment which I have investigated. 



-(2) 



(1) 



(3) 



B C D E F G H 



Figure III. — Intensity curves of the spectrum of Antimony Oxide. 

 (1) Reflexion spectrum of a thin film. (2) Reflexion of a thick film. 

 (3) Absorption spectrum of a thick film. 



