84 THE WANDERINGS OF THE NORTH POLE. 



the equator. It is (jiiite true that as the pole is implied by these circles 

 rather tlian directly marked by tliem, the measurement of the altitude 

 can not be effected (piite directly. The actual process is to take the 

 Pole star, or some one of the other circumpolar stars, and to measure 

 the greatest height to which it ascends above the horizon, and the 

 lowest altitude to which it declines about twelve hours later. The 

 former of these is as much above the pole as the latter is below it, so 

 between them we are able to ascertain the altitude of the pole with a 

 high degree of accuracy. It is true that in a fixed observatory .such 

 as Greenwich there is no visible sea horizon, and even if there were 

 it would not provide so excellent a method as is offered by tlie equiv- 

 alent process of first observing the star directly and then observing 

 its retlection from a dish of mercury. In this way the altitude of the 

 star above the horizon is deteruiiued with the utmost precision. 



The practical astronomer will however remember that of couise he 

 has to attend to the effects of atmospheric refraction, which invariably 

 shows a star higher up than it ought to be. This can be allowed for, 

 and in this way the latitude of the observatory is ascertained with all 

 needful accuracy. When the highest degree of i)recision is sought for, 

 and it IS only observations with a very high degree of precision which 

 are available for our present purpose, a considerable number of stars 

 liave to be employed, and very many observations have to be taken at, 

 different seasons of the year so as to eliminate, as far as possible, all 

 sources of casual error. When, however, due attention has been paid 

 to those precautions which the experience of astronomers suggests, the 

 result that is obtained is characterized by extraordinary precision. 

 How great that precision may be I must endeavor to explan. The lati- 

 tude o± every important observatory is obtained from a larg'e number 

 of observations, and it would be unlikely that it was more than one or 

 two-tenths of a second different fi'om the actual mean value. Now a 

 tenth of a second on the surface of the earth corresponds to a distance 

 of about 10 feet, and this means that the latitude of the observatory, or, 

 as we must now speak very precisely, the latitude of the center of the 

 meridian circle in the observatory, is known to a degree of precision 

 represented by a few paces. It will thus be seen that with the accu- 

 racy attainable in our modern observations, it would often be an appre- 

 ciable blunder to mistake the latitude of one wall of the observatory for 

 that of the opposite wall ; in other words, we know accurately to within 

 the tenth of a second, or within not much more than the tenth of a sec- 

 ond, the distance from the center of the transit circle at Greenwich, 

 down to the earth's equator. But, of course, the distance from the 

 pole to the e(]uator is 00°, and this being so it follows that the distance 

 from the north pole of the earth to the center of the transit circle at 

 Greenwich Observatory has been accurately ascertained to within one 

 or two-tenths of a second. If any change took place in the distance 

 between the pole and the meridian circle at Greenwich, then it must be 



