JANUAEY 12, 190O.] 



SCIENCE. 



73 



In connection with that portion touching 

 upon the Meridian Circle, Professor Keeler 

 states that during the year ending last Septem- 

 her, and upon 106 nights, 6000 observations 

 were made with the Meridian Circle. 



Now the question I would submit, and upon 

 which I would request information, does Pro- 

 fessor Keeler wish it understood that 6000 

 star places were observed in that time, or are 

 these numbers of observations made up of the 

 determination of the right ascension, declina- 

 tion, nadir point, collimation, level and azimuth, 

 each individual determination of these quanti- 

 ties to be counted as one observation ? 



The above figures give as a nightly average 57 

 observations, and from my experience four or 

 five observations per hour — I mean a complete 

 determination in both right ascension and decli- 

 nation, when one is not working with an assist- 

 ant and not in zone work — is about the limit. 



When it is taken into consideration that one 

 observer sets his circle, reads four microscopes, 

 observes nine or eleven transits, makes two or 

 more bisections in zenith distance, and records 

 all these, reads his level at least once every 

 hour, observes his collimation twice in an even- 

 ning's work, an average of 57 observations per 

 night is almost, if not quite, unrivaled. 



But, as I said before, perhaps what Professor 

 Keeler wished to convey by the word ' observa- 

 tions ' is not what I have construed it, a com- 

 plete determination of the two coordinates of 

 the star place, but may contain two, three or 

 four quantities, which "he calls observations. 

 Geo. a. Hill. 

 Naval Observatory, Washington, D. C. 



note on the foregoing letter by professor 



HILL. 



In the part of my report to which Professor 

 Hill refers in his letter, one observation means 

 one complete determination of both coordinates 

 of a star. A complete observation of the nadir 

 (zenith) point and level is also, in accordance 

 with the usual custom, counted as one observa- 

 tion. Collimation and flexure determinations 

 and mire readings have not been included. 



A reference to our records for the year cov- 

 ered by my report shows that the average num- 

 ber of stars, completely observed in both coor- 



dinates during this period by Professor Tucker, 

 was fifty per night. With an assistant reading 

 the microscopes the average number was sixty- 

 two per night of from four to five working 

 hours. The observations, as shown by their 

 probable errors on complete reduction, are of 

 the highest order of precision. 



Doubtless this is quick work, but I believe 

 that it is by no means of unprecedented rapidity. 

 It is moreover obvious that a comparison of the 

 work of different instruments, on the basis of 

 such figures as those given by Professor Hill, 

 may be quite misleading, since the rapidity with 

 which observations can be made depends largely 

 on the character of the work which is being done. 

 With a full list like that of Blr. Tucker's during 

 the past year, the stars culminate more rapidly 

 than they can be observed, so that the list has 

 to be gone over several times. The rate of 

 observation then depends upon the observer's 

 quickness and skill. With a list which con- 

 tains many gaps, stars have to be waited for, 

 and the rate depends upon the list alone. 



James E. Keeler. 



dark lightning. 



To THE Editor of Science : My attention 

 was drawn to Mr. Clayden's work by an article 

 in Nature in which reference was made to a 

 communication in one of the photographic 

 journals. The note in the Philosophical Maga- 

 zine I had somehow overlooked. 



Mr. Clayden in his letter states that he was un- 

 able to obtain any results with the calcium 

 light or with sunlight, and suggests that there 

 may exist some difference between light from 

 such a source and a source whose excitement is 

 electrical, and that it is not safe to assume that 

 the time factor is the only one, until the image 

 of some non-electrical source has been reversed. 

 I cannot see much difference between the cal- 

 cium light and the arc, for in both we are deal- 

 ing with an incandescent solid. To settle the 

 matter definitely I have repeated the experi- 

 ment with the revolving disc, using a calcium 

 light, and obtained perfect reversed images of 

 the slit on the first trial. Mr. Clayden's failure 

 to get reversal with sources other than the spark 

 was due, I imagine, to a too long exposure. The 

 duration must be something less than 1/15000 



