34 On Errors in the Nautical Almanac. 



and expense of forming more shafts than one to a mine ; at 

 least, for the purpose of ventilation ; for this migiit be effected 

 by means of a tube carried down in one side or angle of a single 

 shaft, and extending to the various parts of the excavation be- 

 low; and in extensive mines, several fans might be used with 

 combined effect. 



Mr, Editor, I am aware that the fan has been used to llntu 

 down a shaft a short way, but I am not aware that it has been 

 employed in the above mode for extracting the foul gases from 

 mines. Nor do I know that the vessels inverted in water have 

 been used or proposed as a blowing apparatus. Tiie powers and 

 capabilities of the fan will appear to anv one who has seen Mr. 

 Sadler inflating his exhibition balloon. 



The above ideas, on employing the fan, were lately read to the 

 Glasgow Philosophical Society at two of their weekly meetings. 

 The members present recommended that tiiey should be given 

 to the public through some of the scientific journals. 

 I am, sir, 



Your obedient servant, 

 Glasgow, lethjuly 1816. JaMES WatT, M.D. 



XII. On Errors in the Nautical Almanac, <^c. 



To Mr. Tiiloch. 



Sir, — jyiucH has been lately said respecting the comparative 

 meritsof the Nautical Almanac, and the Connaissavce des Terns : 

 and attempts have been made to show that the computers of the 

 latter work have borrowed very considerably from the former. 

 There are two facts, however, which I have recently discovered, 

 that will (I think) evidently show that the computations in those 

 works are carried on independently of each other. These facts 

 relate to Jupiter's satellites. 



The conjigurations of those satellites, as set down in the Nau- 

 tical Almanac for the last month (June), are almost all of them 

 wrong ; as I found from actual observation, and as may be readily 

 proved by calculation: many of them likewise are incorrectly 

 stated for the ensuing month of August. In the Con?iaissance 

 des Terns, however, the positions are truly stated. 



On the other hand, the eclipses of those satellites are all cor- 

 rectly computed in the Nautical Almanac: but, in the CoH' 

 naissance des Terns the eclipses of tlie^r^^ satellite are set down 

 very incorrectly; there being a constant error, which sometimes 

 amounts to six minutes of time, either in excess or defect. This 

 mistake runs through the whole of the years 1815, 1816 and 



1817; 



