186 



Performance of Borne Timekeepers. 



to lose their trade. According to a notice in the Horological 

 Journal, Australia is England's best customer for watches, 

 still the manufacture has lately fallen off to such an extent, 

 that England does not now export one watch per head of her 

 watchmaking population. The competition the English have 

 to contend against is shown in the case of watch g, whose 

 rates were given before, which came into my possession in a 

 very ordinary way. On passing down Swanston-street I saw 

 it in Mr. Joseph's window; it was simply marked " Waltham," 

 with the price, four guineas, attached. I went in on the 

 afternoon of December 4th, 1878, purchased it, and took it 

 home with me. I compared it with our standard clock, and 

 got the following results : — 



Date. 



Error. 



Date. 



Error. 



d. b, 



m. 



s. 





d. h. 



m. 



s. 



Dec. 4. 5 



26 



12 slow 



Dec 



8. 21 







13*5 slow 



„ 4. 21 







13 „ 



5? 



9. 1 







12-5 „ 



„ 5. 21 







12-0 „ 



JJ 



9. 5 



5 



11-2 „ 



i, 6. 1 







13-0 „ 



>) 



9. 21 



3 



11-3 „ 



„ 6. 4 







12-5 „ 



>> 



10. 1 







10-5 „ 



„ 6. 21 







12-5 „ 



?) 



10. 4 







10-5 „ 



„ 7. 7 



23 



12-0 „ 



J? 



10. 9 



35 



10-5 „ 



„ 7. 20 



31 



12-5 „ 



5? 



10. 22 



12 



12-0 „ 



I was fairly astonished at such a result. On the last 

 mentioned day I left Melbourne for New Zealand. During the 

 voyage I used the watch as a chronometer for finding the 

 longitude, as I had carried a small sextant with me for my 

 amusement. On reaching Dunedin I found that in seven 

 days the watch had altered its error about half a minute, 

 having lost more in the much colder temperature at sea, than 

 in Melbourne. I subsequently found by direct experiment 

 that it was over-compensated to the extent of gaining twelve 

 seconds a day for an increase of ten degrees Fahr. About a 

 year ago I had the screws shifted, and the compensation is 

 now nearly perfect. I still occasionally wear it, and it goes 

 as well as ever; the improvement in the compensation being 

 partly neutralised by an increase in tl>e position error. Mr. 

 Dent says that an adjusted going barrel lever may be 

 expected, with an ordinary wearer, to have a daily variation 

 of rate from 2 to 4 seconds. The above watch did not 



