of the Acceleration of Gravity for Tokio, Japan. 45 



ought to regard "as dust thrown into the pupils' eyes, to 

 prevent their attending to such well-known reductions as 

 those for arc, buoyancy, resistance, &c." It is not quite clear 

 from this whether (a) he means that these corrections were not 

 made in the experiments with the Kater's pendulums about 

 which he goes on to speak; or whether he means (b) that 

 they were not made in the experiments with our long wire 

 pendulum, from which alone our value of g was determined ; 

 or whether, lastly, he desires to imply (c) that, being unac- 

 quainted with the method of applying such simple corrections, 

 and being of course also unacquainted with any text-book 

 in which an account of such corrections could be found, we 

 slurred over the matter and intentionally abstracted the stu- 

 dents' attention by "abstruse investigations of conditions 

 which, almost obviously, are unimportant." 



(a) We reply that, as we do not give in our paper, or in any 

 way use for the numbers there mentioned, any of the experi- 

 mental results obtained with the Kater's pendulum, it appeared 

 to us unnecessary to describe the corrections proper to be em- 

 ployed when using the reversible pendulum. It may, how- 

 ever, now be mentioned that, after the knife-edges of the 

 reversible pendulum had been finally adjusted so that the pen- 

 dulum beat seconds either end up, it was always found that the 

 distance between them was slightly greater than one metre. 

 Now the reductions for arc, for buoyancy, and for resistance 

 would, when applied, have each separately slightly increased 

 this apparent length of the seconds-pendulum. Consequently, 

 as on no part of the earth's surface is the length of the equi- 

 valent simple seconds-pendulum greater than one metre, it 

 follows that, whatever may have been the cause for our results 

 with the reversible pendulums not being satisfactory, whether 

 it was due to our own inexperience, or whether it arose from 

 defective workmanship in the instrument itself, want of abso- 

 lute parallelism of the knife-edges, &c, it certainly could not 

 have been due, as Major Herschel appears to imply, to any 

 want of " attending to such well-known reductions as those 

 for arc, buoyancy, resistance, &c." 



(b) and (c). But even were this not so, would Major Her- 

 schel be justified in assuming that we made no corrections 

 when using the Kater's pendulums, seeing that we have made 

 and given in our paper the results of the much smaller correc- 

 tions for arc, for buoyancy, and for resistance in the case of 

 the long pendulum, in addition to those other corrections 

 which your correspondent is pleased to call " abstruse and 

 almost obviously unimportant," but which have appeared to 

 Sir William Thomson of sufficient interest to induce him, 



