62 Professor Forbes' s Experiments on 



9. My method of observing was to keep the chronometer 

 at the ear till the instant that the termination of a vibration 

 was observed, then to count five beats of the balance (corre- 

 sponding to two seconds), which affords time to bring the 

 dial-plate into view, and the seconds entered in the Table are 

 those read off at that time, namely two seconds later than the 

 absolute times. Thus the impracticable attempt to observe 

 two things at once by the eye is avoided. For this and some 

 other suggestions, I am indebted to my friend Captain P. P. 

 King, R.N. The observations are registered in lithographed 

 forms bound into volumes. 



10. In the choice of stations I have been extremely parti- 

 cular, often at great personal inconvenience*. Places remote 

 from any trace of habitation have most usually been selected, 

 and in no case have intensity observations been made in a 

 house. The specialties of the sites will be noticed in the fol- 

 lowing Tables. I have invariably removed all masses of iron 

 from my person ; and in my later experiments even took 

 the precaution of carrying thin shoes, in order that the heavily 

 nailed ones which I usually wear, might be removed to a di- 

 stance. The chronometer, too, has generally been held at 

 some distance from the apparatus. But some direct experi- 

 ments lead me to believe that the influence of the two last- 

 mentioned sources of error is insensible. 



§ 2. Corrections applicable to the Observations. 



1 1 . When the mean of seven values of 100 or of 300 vibra- 

 tions has been taken, as above explained, a variety of import- 

 ant corrections remain to be applied. 



12. I. Rate of Chronometer. — The following rule due to 

 Professor Hansteen is simple and accurate : — " The logarithm 

 to five decimal places of the observed time is taken, unity is to 

 be added to the fifth decimal place for every two seconds per 

 diem that the watch goes slow; and unity subtracted for every 

 two seconds that the watch goes fast." The demonstration is 

 too simple to require notice. The following is a table of cor- 

 rections : — 



Table I. 



Log. Additive. Log. Additive. 



Rate + sec 0-00000 Rate — sec 0-00000 



2 0-99999 2 0-00001 



4 0-99998 4 000002 



6 0-99997 6 0*00003 



8 0-99996 8 0-00004 



10 0-99995 10 0-00005 



* None but those who have been engaged in observations of the very 

 same description, where the eye, the ear, and the memory are all actively 

 employed, can have an idea of the difficulty of always finding sites free from 

 the interruptions of curiosity, or natural obstacles. 



