G. K. Gilbert — Congress of Geologists. 4A7 



each period, not a certain portion or area of the spectrum, but 

 a specific color defined by its position in the spectrum. This 

 color scale will also apply to the geology of Europe. I pro- 

 pose, third, that the students of each geologic district shall 

 assign to the stratigraphic systems of that district a set of pris- 

 matic colors so selected from the spectrum as to properly repre- 

 sent the relation of each system to the time scale, provided 

 that relation is approximately known. Under this rule a sys- 

 tem corresponding partly with the Cretaceous and partly with 

 the Jurassic will receive a prismatic color intermediate between 

 those assigned to the Cretaceal and Jural divisions of the time 

 scale. I propose, fourth, that systems whose relations to the 

 standard time scale are not even approximately known be 

 given tentative positions in the time scale and assigned the cor- 

 responding colors; and that such provisional colors be distin- 

 guished by a special device. 



Of this device I will speak later, but before we leave this 

 part of the subject, the capability of the plan to express the 

 facts should be more clearly characterized. Continuous geo- 

 logic time being equated with the continuous spectral band of 

 light, each period is theoretically equated with a segment of 

 that band including all the hues between certain limits. But, 

 practically, the period is represented in the color scale only by 

 the central hue of the segment, and there is nothing in the 

 nature of this hue to indicate the length of the segment. 

 Similarly each local system is represented only by the hue cor- 

 responding to the middle of the equivalent period, considered 

 as a part of the continuous time-scale, and this hue gives no 

 information as to the magnitude of the system or the duration 

 of the corresponding period. When a non-European system 

 is represented on a map with the Devonal color, all that is ex- 

 pressed is that the middle of its period coincides with the 

 middle of the Devonal period ; the whole period may equal 

 the Devonal or may be shorter or may be longer. With this 

 limitation the scheme is able to express the exact facts, or the 

 exact state of opinion, in regard to correlation. 



I propose, fifth, that the subdivisions of systems be repre- 

 sented, if their number is small, by distinct tones of the hue 

 assigned to the system, and if their number is great, by mono- 

 chromatic textures. It having been provided that systems 

 shall be distinguished by means of hues, it is now provided 

 that hues shall have no other function. This secures the in- 

 tegrity of the distinction between systems, whatever the 

 minuteness of subdivision. 



The idea of using the spectral colors in their proper order is 

 not novel. It has entered into half the plans submitted to the 

 Congress, but each author has introduced other colors also, or 



