MAGNETIC PRECAUTIONS 43 



tail end of the shaft was made hollow, and inside it was placed 

 an inner shaft ; the outer shaft fitted into the boss of the 

 screw on a taper ; inside the boss beyond this taper was a 

 large nut in which the inner shaft could engage ; the outer 

 shaft and the screw were kept in close connection by the inner 

 shaft and nut, and therefore there was no loose connection to 

 jar. To disconnect the screw, a small section of the main 

 shaft, in front of the tail shaft, could be lifted bodily, the 

 inner tail shaft could then be turned and freed from the nut, 

 when both inner and outer shafts could be withdrawn together, 

 and the screw was free for lifting. This fitting was naturally 

 expensive, but it is certainly the most efficient that has been 

 devised for a lifting propeller. 



In the profile drawing which is reproduced, on the middle 

 of the upper deck will be seen a deck-house marked ' Magnetic 

 Observatory ' ; this was an important place, both in the 

 building and in the subsequent work of the * Discovery.' I 

 have already given reason to show why the greatest stress was 

 laid on the accuracy of our magnetic observations, and it will 

 be clear that accurate magnetic observations cannot be taken 

 in a place closely surrounded with iron. The enthusiasm of the 

 magnetic experts on the Ship Committee had at first led them 

 to request that there should be no iron or steel at all in the 

 ' Discovery,' and when it was pointed out that this could 

 scarcely be, they demanded the exclusion of the metals from 

 the vicinity of the magnetic observatory. At last a compromise 

 was arrived at, which stipulated that no magnetic materials 

 should be employed within thirty feet of the observatory. It 

 is difficult to realise what immense trouble and expense this 

 decision involved. This thirty-foot circle swept round, down 

 by the foremast, under the bottom of the ship, and up in 

 front of the mainmast ; everything within this radius had to 

 be made of brass or some other non-magnetic material, and 

 when all the fastenings of the hull and all the fittings and 

 furniture of the ship are considered, some idea may be gathered 

 of the difficulty ; even much of the rigging, which would 

 ordinarily have been of wire, had to be made of hemp, of a 



