462 Bunsen and Eoscoe's Pendulum Actinometer. 



been abandoned in favour of clockwork, which, after many 

 trials, has been adapted so as to give and maintain the required 



Fig. 4. 







speed at a uniform rate, for periods well within the limits 

 required for our purpose. 



The spindle of one of the rollers is attached to an 8-day 

 English clock from which the escapement has been removed, 

 the change of speed ordinarily occurring when a clock is 

 running down under such circumstances being compensated 

 for by means of a fusee adjustment. 



The following measurements were made to determine to 

 what extent the clock, thus modified, could be relied uoon to 

 give a constant speed to the shutter at different stages of the 

 unwinding of the spring. 



Four sets of ten consecutive readings were taken of the 

 time of opening of the shutter. Two series of measurements 

 were made. 



These numbers show that during forty revolutions of the 

 shutter the variation in the time of exposure is exceedingly 

 small, and it is obvious that by using the same portion of the 

 spring in each case 10 or 15 exposures may be made where 

 the time is practically constant*. 



* The fact that it takes a slightly shorter time to open the shutter 

 when the spring has partly run down, as is seen from the above readings, 

 must he due, we think, to the fusee not being turned quite accurately. 



