,310 



PROFESSOR L. BECKER ON THE DISTRIBUTION OF 



It can be shown from the observed durations that the clock was off-beat at the 

 Observatory, and probably it was so, too, at the eclipse, and possibly in the opposite 

 way. This would not affect the relative duration of the first five contacts, as they all 

 lie between an uneven and even second, but their errors would appear relative to the 

 long exposures. 



The sequence of events governed by the clock takes place at the moments of time 

 shown in the following table, where the numerals denote the seconds elapsed from 

 second 0, when the pendulum is started : 



r designates the ratio of the focal-length and the diameter of a lens, which has the 

 same area as the lens reduced by the screen. 



There are several points in the design of the apparatus which have proved 

 unsatisfactory. The shutter must have a smaller moment of inertia, and its motion 

 should be recorded on a chronograph ; the plate-holder ought to run on wheels 

 instead of sliding on a rod. The mutual distances of the pictures ought to be, say, 

 four solar diameters, and, especially, the side of the square opening in front of the 

 plate must be twice as great as the distance between the pictures [see 5 (/), (g), (/<)]. 

 One of the pictures (not the last) must be 8 diameters from its neighbours [see 4 (c)]. 

 The screen for cutting down the aperture of the lens ought not to contain a series of 

 small openings, but have a central opening and an annular opening, whose diameter is 

 about two-thirds of that of the lens (see Appendix I, p. 332). 



