Hartshorne.] ^i-U [April 21, 



through exliaustion from special use; wliile, on the theory which I am 

 about to set forth, the mode of causation of the secondary as well as of the 

 primary spectrum appears quite clear. 



It remains, then, to attempt a definite answer to the question, what ex- 

 planation, conforming with all these facts, can be substituted for that which 

 I have ventured to pronounce unsatisfactorj'. I mako an endeavor in this 

 direction with diitidence; not possessing so precise an acquaintance with 

 physics as might enable me to deal as an expert with so diificult a subject. 



Allusion has been made above to four kinds of phenomena; Avhich may 

 be thus brielly named, en resume: 



a. Those of ordinary color spectra, primary and secondary, seen with 

 moderate light. 



b. Color shadows thrown under transmitted light. 



c. Ocer-tints and comj)lementary shades, seen through thin paper, on a 

 brightly colored ground. 



d. Solar color spectra, positive and complementary by turns, according 

 to whether the eyes are closed, or opened towards a white ground. 



These all appear to me to have a common character, and to require an 

 essentially identical explanation. 



Take, first, the ordinary color spectra, such as may be obtained with 

 good lamplight. Looking at a red object for a few moments, one then turns 

 the eyes to a white surface; a green spectrum is seen. Why is this? Be- 

 cause, to use the simplest and least hypothetical phraseology, the eyes are 

 charged, saturated with red light; and this, having a certain strength, is 

 neutralized by the red ra^-s in light reflected from the white surface, so that 

 only the remaining, comjilementary, green rays of that light affect the 

 sight. Translating these expressions into language in accordance with the 

 undulatory theory, I would say that, when a brightly colored object is 

 looked at, those rods and cones, or minute retinal nerve-elements, which 

 respond in vibration to the luminous ether-waves of the color reflected to 

 the eyes, are excited to moli(m thereby; and by "irradiation," or communi- 

 cation of vibrations, all retinal elements which have the same period of 

 vibration are made to partake, in some degree, of this movement. Then 

 when, turning from the colored object, wliite light, consisting of all the 

 color-rays or waves together, impinges ujion the eyes, those ether-waves of 

 the white light which belong to the color first acting on the retinal nerve- 

 elements, interfere with, and for the time relatively diminish or annul W\e 

 special vibralicms already i)roduced in the retina; leaving the other waves 

 of white light to take eflect upon the retinal elements which respond to or 

 "resonate" with them, so that the comjileinentary color only is seen. 



Very \'>ro\nMy relative diminution, rather than total arrest, of the special 

 retinal vibrations, is what occurs. All our perceptions of light and color 

 are, to a considerable extent, dependent on the relative intensity of light 

 from different sources or of different kinds. When, then, in a beam of 

 white light, & portion o? cvxhun color rays is, in effect, arrested by previously 

 existing retinal vibrations of the same period, although the remaining rays 



