190 



THE LAKE REGIONS OF CENTRAL AFRICA. 



time by a sixpenny sun-dial. Before the first fortnight 

 after our second landing in Africa had elapsed, all these 

 instruments, notwithstanding the time t and trouble de- 

 voted to them by my companion, at Zanzibar, failed in 

 their ratings and became useless for chronometric lon- 

 gitudes. Two of them (Ed. Baker, London, No. 863, 

 and Barraud, London, No. jfy), stopped without ap- 

 parent reason. A third, a first-rate article (Parkinson 

 and Frodsham, No. 2955), issued to me from the Royal 

 Observatory Greenwich, at the kind suggestion of Capt. 

 Belcher, of the Admiralty, had its glass broken and its 

 second-hand lost by the blunderer Gaetano : we re- 

 medied that evil by counting the ticks without other 

 trouble than that caused by the odd number, — 5 to 2 

 seconds. This instrument also summarily struck work 

 on the 9th November, 1858, the day before we intended 

 to have "made a night of it" at Jiwe la Mkoa. This may 

 serve as a warning for future travellers to avoid instru- 

 ments so delicate that a jolt will disorder them — the 

 hair-spring of the lever watch w T as broken by my com- 

 panion in jumping out of a canoe — and which no one but 

 a professional can attempt to repair. A box chrono- 

 meter carried in a "petarah" by a pole swung between 

 two men so as to preserve its horizontally, might out- 

 last the pocket-instruments, yet we read in Capt. Owens 

 celebrated survey of the African coasts, that out of 

 nine not one kept rate without fluctuations. The best 

 plan would be to purchase half-a-dozen sound second- 

 hand watches, carefully inspected and cleaned, and to 

 use one at a time; if gold-mounted, they would form 

 acceptable presents to the Arabs, and ultimately would 

 prove economical by obviating the necessity of parting 

 with more valuable articles. 



The break-down of the last chronometer disheartened 



