^{page 155 note *} A full account of these observations was published in the Edinburgh Journal of Science for January and April 1828.

^{page 161 note *} I do not except Mr Daniell's Observations for reasons assigned below, §. 9.

^{page 161 note †} Humboldt, , Relation Historique, 4to edition, tom. iii. p. 306Google Scholar.

^{page 161 note ‡} Astronomische Nachrichten, 1823, No. 26.

^{page 161 note §} Philosophical Transactions for 1827, &c.

^{page 161 note ∥} See M. Bouvard's Memoir in the Memoires de l'Institut for 1824.

^{page 163 note *} For part of those for 1830, I am indebted to Mr Hudson, the observer, whose zeal in the examination of this phenomenon promises soon to afford us new and valuable data.

^{page 163 note †} Bibliotheque Universelle, xx. 246.

^{page 163 note ‡} Memoires de Tlnstitut, 1812.

^{page 164 note *} Humboldt, Relation Historique, iii. p. 302

^{page 164 note †} Ibid. p. 298.

^{page 164 note ‡} Philosophical Transactions, 1828.

^{page 167 note *} A similar fact with regard to Northern Europe has been suggested by some observations made at the apartments of the Royal Society of London, in a paper by Mr Lubbock, V.P.R.S. in the Phil. Trans. for 1831, of which the author has been kind enough to favour me with a copy since this paper was read.

^{page 171 note *} Des Lignes Isothermes. Mem. d'Arceuil, tom. iii. Table.

^{page 172 note *} Indeed there seems much reason to doubt whether the period of the day alluded to has a mean temperature below 0° cent. If it has not, M. Bouvard's formula would totally fail, as the equatorial oscillation would come out with a wrong sign.

^{page 174 note *} In 1829, the mean temperature of the seasons, distributed as above, was − 1°.5 R. + 4°.8 R. + 0°.5 R., and – 4°.6 R. In 1830, it was colder, as follows: − 0°.1 R. + 4°.5 R., − 0°.7 R., and − 7°.7 R. respectively.

^{page 176 note *} The principal deviation which I have remarked is in the extensive series of M. Marqué Victor, at Toulouse, where the observations were made many times a-day, and probably had M. Bouvard's formula of reduction for hours applied to it. However, it seems rather unaccountable how any formula of this kind should diminish instead of increase the amount of oscillation, which it actually does in this case, unless there be some numerical mistake, the amount derived by M. Marque Victor (and which I have verified as far as the combination of the annual means is concerned) being l^mm.20; according to M. Bouvard it is l^mm.00. As, however, M. Bouvard treats this as the best determined point in Europe, it is probable that he had carefully examined the tables in the Memoires de l'Academie de Toulouse, the later volumes of which I have not been fortunate enough to see.

^{page 183 note *} Thus Königsberg, though in a lower latitude, appears to have a smaller oscillation (see § 6.) than Edinburgh; but then the temperature appears to be only 43° F. (Astronomische Nachrichten, Feb. 1823), while that of Edinburgh is 47°. It must be hardly necessary here to observe, that, supposing the mean temperature of a place known in a function of the latitude, my formula admits of a direct corncomparison with the temperature (t) at the level of the sea. Thus, if Dr Brewster's formula of t = 81.5 cos θ be employed, my formula becomes ; a being a new coefficient.

^{page 184 note *} Biot, Notice des opérations enterprises pour déterminer la figure de la terre.

^{page 184 note †} The results are published in the Society's Transactions, Vol. X.

^{page 185 note *} The abstracts of Captain King's observations are about to be published in the proceedings of the Royal Geographical Society, vol. I.