552 Mr. P. L. Gray on the Minimum 



resting on the slate block below the strip, as in the calibration 

 experiments; thus (natural size) : — 



The angular dimensions of the surface of platinum, as seen in 

 any experiment, were therefore ; — 



Apparent length = 3° 49' approximately ; 

 „ width =1° 54' „ 

 so that the apparent area subtended was about 36 times that 

 of the full moon. 



The current by which the strip was heated ran through a 

 variable carbon resistance, the handle of which was within 

 convenient reach of the observer as he sat with his head under 

 the black cloth. He could thus alter the temperature of the 

 platinum until it was on the very verge of invisibility, a very 

 small fraction of a turn being then sufficient to produce utter 

 darkness where before the area of faint light had been. A 

 contact-breaker was also within convenient reach, so that the 

 current could be broken or made at pleasure, and the objective 

 reality of the faint luminosity at the limiting-point thus de- 

 monstrated. When he was satisfied that the limiting-point 

 had been reached the hinged end of the box was opened, a 

 beam of light sent to the mirror connected with the strip, and 

 the deflexion, giving the temperature, read on the scale. The 

 possible error in the estimation of the absolute value of the 

 temperature may be taken as certainly not more than 2°. 



Discussion of the Results. 



My first idea, in putting down final results, was to take the 

 mean of a large number of observations as expressing the 

 required minimum temperature ; but there soon appeared to 

 be too much variation in individual cases, both for the same 

 eyes at different times and for different people's eyes, for this 

 mean to be of any value. It seems better, therefore, to put 

 down the determinations in some detail, with such remarks as 

 may help towards some general conclusions afterwards. 



