NEAR FORT ST. GEORGE. 517 



chance ; and it was this consideration which led me to 

 examine whether I could, not discover some law by 

 which it was governed. 



p. For this purpose I gathered all the observations 

 which I could collect, and from as many different places 

 in the Peninsula as I could, provided they were obtained 

 in sufficient luimbers at each place. These being 

 arranged in the order of the declinations, and the mean 

 results of northern and southern zenith distances being 

 taken separately, I noticed invariably (though in une- 

 qual degrees) a small excess in the northern, and defect 

 in the southern sets. 



10. An abstract of these deviations is given in Table 

 IV. and the reader will do well to refer to it in order to 

 judge of the consistency of the preceding remark, and 

 of the solidity of what I have further to say on the sub- 

 ject. 



11. This exposition being sufficient to shew the 

 tendency of observations taken north of the zenith to 

 give too great a latitude, and the contrary of the south-- 

 ern ones, I shall now endeavour to account for this cir- 

 cumstance as follows. 



12. Let Z f Plate V\. Fig. l) be the true zenith of 

 any place. E D, and E d, the declinations of any two 

 stars, one north, the other south, and nearly at equal 

 distances from the said zenith. By the present experi- 

 ments, if we use the declination of i), the latitude will 

 fall somewhere in L: but if we use the declination of 

 d, then it will fall on the opposite side, somewhere in 

 /. Therefore the sum of the zenith distances D L\-ld, 



Ll3 



