174 PROCEEDINGS OP THE AMERICAN ACADEMY 



the "latitude of the axis" and the "half extent in latitude" just ex- 

 phiined, by an increase of thirty degrees in the elongation. Here also 

 only those observations have been used which afforded complete deter- 

 minations, and the results will differ from those directly obtainable by 

 comparison of the quantities found in Tables XVI. and XVII. This 

 restriction of the observations, of course, often makes the number 

 which can be used too small to permit any inference from the result 

 until it lias been combined with others to form a larger group. The 

 data of Table XVIII. are given under the headings " Dis[)lacement " 

 and " Contraction," respectively, for stated changes of elongation. 

 Under " Displacement," the negative results expressed by Italic figures 

 indicate that the axis lies firther south at tiie greater elongation. 

 Under " Contraction," tlie diminution of the " half extent in latitude" 

 is expressed as a positive quantity, and a negative result would mean 

 that this extent increased, instead of diminishing, with the elongation. 

 In discussing the observations of Table XI., those made by Eylert 

 and by Neumayer, owing to their small number, can be reduced sepa- 

 rately with little trouble. It seemed best, however, to furni>)h an 

 intermediate step in the reduction of those made by Ileis and Weber, 

 (as has already been done for those of Jones by Tables XIII. to XVIII. 

 inclusive). Tliey have accordingly been arranged according to the 

 calendar montiis, and the mean results thus obtained are given in Tables 

 XIX. to XXII. inclusive. The form of Table XIX. is nearly the 

 same as that of Table XIII. As only one observation was made on a 

 single d;iy, the column "No. of Days" is omitted; and the column of 

 months does not proceed in chronological order, as in Table XIII., 

 but combines the observations of dilTerent years. Since Heis and 

 Weber sometimes observed an " inner cone," the mean elongation of 

 its vertex is here given in a space which would otherwise be vacant ; 

 the spparate observations are too few to require a place in Table XI. 

 Table XX. gives for Ileis and Weber results corresponding to those 

 of Tables XIV. and XV. for Jones. Tables XXI. and XXII. in like 

 manner correspond to Tables XVI., XVII., and XVIII., with one dif- 

 ference. From tile form in which Ileis gives the observations, it is 

 sometimes practicable to find llic; latitude of the axis at elongations for 

 which the latitudes of the boundaries cannot so well l)c estimated. 

 Tiiis allows the number of observations of the axi.s, in Table XXL, 

 occasionally to exceed that of the observations of " half extent." 

 Similarly, in Table XXII., the numbers for "Displacement" and for 

 "Contraction" sometimes differ, and are therefore given in separate 

 columns, instead of together, as in Table XVIII. 



