8 



The iiltsoliilc (1i-U-niiiniiti()ii oC tlic fun-c c.f ;;nivi(y iit any |«iiiit <•• any 

 gri'iit (U'mtr of |>nrisii>n is ii miUU-r involving; iimiiy ililliciillii« iiml llii« is 

 f8|M.rially tnii- iiiulcr circiimstjiiici's in wliiili tlic liuililifs for dniip^ tin- work 

 arc ceitaiiily imt of tint lx'«l. AsicK: from llif cxiiiriiMcntal «lilliciilticx, tlicrcun- 

 iiniiicruii8 Boiirtvs of iMtssibli', imUixl pniljulilc, error which can only Ik- iiivc«ti- 

 f;atoil ii\u\ properly disposed of under e.vceplioimlly fiivonible conditioiiH. 

 rndonliteilly, llierefoiv, more trustworthy results are to Ik; exjKvt«-«! from com 

 pirative detenuiuiitioiiH by m(*iksuring tiie |»eriodic time of the Kiime ]>eiidulum 

 vibnited at different stjitions, the corrections to be applied litivih); lieen carefnily 

 inveslij^ated and its |K.'riod determined at some fundamental station. In acconl- 

 aiice with this view it is pro|H>sed in the immediate fntuif to undertake a 

 careful determination of tin- jieriodic time of such an "invariable p<tidulum" 

 and afterwards to send the siune to l»e vibrated at some jwint in Aineric« or 

 Europe. 



PKKVIOI'S I»CTi:UMINATIONS. 



Up to about the time of the conclusioti of these experimenls I wiw not 

 aware that any previous attempt had Ui-n made to determine the value of the 

 force of gravity at this point. I i«in the arrival of the rhiio!ii>phic;il Ma;;azine 

 lor April 1880, however, it was found to cont4iin a pa|iir by Measr» Ayrton ami 

 I'erry on a "Determination of the Aecelen\tion of (Jravity for Tokio, Jii|«in"— 

 which was ba.'^ed on exi)eriinenfs made by the Authors at the ('olle-^- of Hn<;i- 

 necrinj^ in this city in 1878. An examination of this pa]>er will show that then; 

 are serious objections to the juethod pursued by Messrs. Ayrton and I'erry 

 liesides numerous and fatal crrore committed in the reduction of their results. 



The jienduhnn u.sed by Messi-s. Ayrton and I'erry was neiirly ten nielent 

 in length. Th.'re are serious objo:;tions In tlu uvj of a long pjndiihun. IVtnhi, 

 in his celebrated detenninations made at Paris, used a |ienduliim alx)ut four 

 meters long, but one which ai>pro.ximates in length to a wconds |n'ndulum hiu« 

 l>een almost luiiversally made use of since. The great objection to the use of a 

 long pendulum is the difficulty of measuring it i'« jAacf. Messrs. Ayrton and 

 I'erry measured their inindulum by placing it in a horiy^tntal |Misition, ami 

 stretching it by allowing the end near the ball to hang over a wheel with very 

 little friction. The length was obtained by comparison with a l»ar one meter 

 long, and as this bar must bo placed ten timjs to cover the whole length, it is 

 jilain that any great degree of accuracy must have liecn difficult to obtain, ami 

 this is es|)ecially true when the nie;usuivment of that |Kirtion of the win* which 

 hangs over the wheel is considere«!. Their 'JCith exjieriment wiis made on the 

 25th of January, and the .'»3d on the 2 Ist of February, from which we may infer 



