22 Prof. S. P. Langley 07i Energy/ and Vision. 



required for the perception of very faint-coloured liglits was 

 investigated. The method was an electrical one. There was 

 automatic registration on a chronograph of the instant of 

 exhibition, and determination of the instant of response as 

 the observer pressed a kej. The interval of course includes 

 quite a train of distinct operations. According to Menden- 

 hall (American Journal of Science, [3] vol. ii. p. 156), that 

 portion of the action of brain nerve and muscle which pro- 

 duces the mechanical effect, and which may be called auto- 

 matic, takes place in certainly but little over one tenth of a 

 second. But the sensations which demand a conscious con- 

 centration of the attention, and especially those which require 

 for their registration a decision of the judgment, occupy an 

 interval several times as great. The perception of a light just 

 at the verge of visibility probably involves an exercise of 

 judgment — an answer to the question, " Do I see the light or 

 do 1 not ? " — although the question may not be consciously 

 propounded, and accordingly this kind of perception may be 

 included in that class of combined sensation and mental 

 operation which involves a choice. Professor Mendenhall 

 found for the time required to decide between red and white 

 0"44o sec. and to decide between a circle and a triangle 

 0'494 sec. We have found for the average of over 1000 

 observations of the disappearance or reappearance of a very 

 ftiint light (perhaps 20 times as bright as the faintest per- 

 ceptible), 0*507 sec, but corresponding measurements with a 

 moderately bright spectrum, the light being about 10,000 

 times as intense as that called " very faint," gave 0'242 sec, 

 a number which is intermediate between the times found by 

 Professor Mendenhall for the appearance of a white card 

 (0-292 sec) and that of an electric spark (0*203 sec). We 

 may therefore conclude that distinct vision for a very faint 

 light demands about one half second of time, while the per- 

 ception of light of ordinary brightness requires only about 

 half that interval. It is possible that differences in the 

 rapidity of the perception for lights of different colours might 

 be detected on more exhaustive study, but none have been 

 noted in these experiments other than those which were 

 attributable to the variation of intensity. 



It will be seen that quantitative measurements of the effect 

 upon the eye of different rays whose luminosity varied in the 

 proportion of 200,000 : 1, were actually obtained, and that it 

 would have been possible to considerably exceed these limits, 

 especially when it is considered that the photometric measure- 

 ments were confined to lights of feeble intensity. Since it is 

 possible to look directly at the sun for as short a time as one half 



