148 ANNUAL RECORD OF SCIENCE AND INDUSTRY. 



night by the aid of a Druinmond light, except one which was 

 made by sunlight. Exceptionally still weather was chosen 

 for each experiment. From these observations there results 

 a velocity for light in vacuo of 300,400 kilometers per sec- 

 ond (186,700 English miles), with a probable error of less 

 than ToVo f the whole amount. The solar parallax is di- 

 rectly deducible from Cornu's velocity of light in two w r ays. 

 Thus, combining it with Delambre's value for the equation 

 of light, we find a solar parallax of 8.88", while Bradley's, 

 value of the aberration of light gives a parallax of 8.88", and 

 Struve's value of the aberration constant gives 8.80". M. 

 Corn u in his elegant memoir gives a summary of the values 

 of the sun's parallax as deduced by various methods. The 

 harmony of the results is marvelous when we consider with 

 how minute a quantity we are dealing. The eight values 

 that may be thus deduced range between 8.80" and 8.88"; 

 and it is possible that the transit of Venus may not give a 

 much better determination. 



AUTOMATIC REGISTRATION OF THE CHEMICAL ACTION OF 



LIGHT. 



The measurement of the chemical intensity of the solar 

 light has not yet become a subject of regular meteorological 

 observation, because of the want of a proper instrument. 

 This want is now partly supplied, by a method proposed by 

 Roscoe, who proposes to effect the measurement by the black- 

 ening of a paper saturated with chloride of silver; or, rather, 

 by means of the time required in order that the exposure to 

 the light may bring about a given intensity of shade. A 

 uniformly prepared paper is placed in the apparatus during 

 the previous night, and is, by a mechanical arrangement, 

 hourly exposed during a given interval to the sunlight. In 

 order to estimate correctly the intensity of the solar action, 

 Roscoe arranges the apparatus so that the paper shall, each 

 hour, many times in quick succession, be exposed to the light 

 for from two to thirty seconds. We thus have, at each hour 

 of the day, a complete series of small spots of various tints, 

 and have only to seek that tint which corresponds to the 

 normal to know at once the number of seconds of exposure 

 corresponding to the strength of the sunlight at that time. 

 A series of observations made during the months of May, 



