TRANSACTIONS OF THE SECTIONS. 
21 
loads rae to hope that an equal degree of interest will betaken at this, its second visit, 
m the results of our labours. As regards the purely scientific department, the obser- 
vations have for the most part been printed, and the liberality of the corporation in 
applying the observatory with the magnificent instruments with which these obser- 
vaiims were made, will, I doubt not, be highly appreciated by the distinguished 
v ibiin now in this town. There is, however, one department of the observatory to 
*hich public attention has not been directed so fully us, I think, its importance de- 
mtvi' 4 —I allude to the rating and testing of chronometers 5 this being, moreover, 
the main object for which the observatory was established. In the Memorial from 
“‘■idi I have quoted, it is stated that a captain could not at that time obtain the cor¬ 
net error and rate of his chronometer, in consequence of the time not being accurately 
niuwii. Since the establishment of the observator y, this deficiency has been removed ; 
t ic arrangements for obtaining nnd carrying on’the time ore su complete that no 
"Hailde error can possibly arise in that way. Having accomplished this object, our 
|, Unru>n has been directed to further improvements; and the exquisite arrangements 
' ncli we possess for testing time-keepers, have led us to the discovery of a source 
1 ' trgr » which chronometers employed in the merchant service are subject, the 
",- |r ^ince of attending to which 1 will endeavour to explain. To enable the mariner 
I 1,1 .. ' nn ?itude at sea by the chronometer, he is furnished with its daily gain or 
' . ich is called its rate. The practice almost universally adopted has been to 
; granted that this rate is uniform during the voyage. Chronometers are sup- 
Midv i°, cora P ens ated for the changes of temperature to which they must ncces- 
* JL ’ lx l l0s ’Pd at sea in different latitudes, and captains are instructed to use the 
Plovrd in ilf 8 cl ‘ raates - Now, we are prepared to show that, for chronometers era- 
Pvrsturo t e ® erc hant service, the average change of rate caused by changing the tern- 
r,,!„ / >Dl to b’b° Fabr. is seven secon-ls a day. Hence, if we ascertain the 
Ceratu . f 0met< ' r on 8 ^ orc w «th the most scrupulous degree of accuracy, in a 
brnupms . aru ' ’hi* chronometer be aftei wards taken to sea and exposed to 
' ''itawi'chT 1 000 tni tlie g hort Period of eighteen days, the nccumulated.error on 
nunm,.. n ,.j lnic ' accor ding to the above-named average, will be upwards of two 
mi-;in r ’ 0X1 eiror of two minutes it is stated in the Memorial before referred to 
modi ; n n;n- a . W| ? c . This variation of rate, from change of temperature, differs so 
Hunt in J! time-keepers, that without a trial no idea can be formed of its 
comr„|.. d f ay particular chronometer. In order to show this more clearly, we have 
Bhaueuof woMaofthe observatory, three tables; each table shows the 
I-fftturo to ru eaL ’h oi one hundred chronometers, caused by changing the tern- 
( t , cxtc «t named in the respective headings. Notwithstanding the great 
tiad that . 0Wn ) y tbt ‘ tables to have been produced by change of temperature), 
' orii;i nh i nilie tcen cases out of twenty the original rales return the instant that 
twenty* it j' f,'' m| ] e !'u turp is ^stored. On' an average of about One chronometer m 
chance nr that the rate changes iu the most capricious manner with almost 
? 1 r 'Xt. s of Eiirl» ei ? perature * nntl an examination of our records will show that the 
y placed on .. acllroaom etetB are equally uncertain,—no dependence whatever can 
■£ tho t a ki ”’i ' m perfect time-keepers have necessarily been rejected in 
n . 1 " > were , Uln olliei ' respects the chronometers were taken ill succession 
■ 'fcfiinnSJivi®! lhc obser vatory. They arc arranged in the order of change 
Just tlif. tbe chronometer which gained the most, and ending with that 
in giog the inn, 08 ' * n ' lul> le I » the average change in the daily rate caused by 
U 'Qomcte r fro111 4<) ° to tiO° is r,"'ij7 j taking the two extremes, one 
i? , P*nthii X . o r 1 ] , Ji Ula J r «l Rained 15" 3, and one lost 72"*2 a day, by changing the 
1 1 gaining. »i„ ' ,10 average change of rule of the first ten in the hundred is 
^ V* in th„ aver agc or the second ten is 0”'3 losing; aud the average of the 
-?i ** Dint. „ •* 18 29"*8 losing. There are fifteen chronometers in which the 
'• more than v ,. u 8 losiu K- There are fifteen chronometers m ™ 
. ,rc dian b" and 1 u , n< ^ * US9 than 5" a day, and there are six in winch the loss is 
' jr three clirnn C8S l , <J " a day » so thn't ia quite |>oasiblo for a captain to have 
.'/ , ' aa ’lt[ 1 j 8 loni,it < !, rr } etei ?’, a ^ °** which might place the ship in nearly the same longi- 
I . n " u r* in t Wo ‘! R mi sbt, notwithstanding, by the accumulated errors of the chro- 
1 r * Tables 11 rG ? he wrong to an amount sufficient to place the ship m 
hire f ron| q , • a »d III. show the change of rate caused by changing the tem- 
10 80 and from 50° to 80° respectively ; the variations, it will be 
