that no experiments are valuable, unless the measures are most minute, and the accordance amongst them most perfect. It may, perhaps, be of some use to show, that even with large instruments, and most practised observers, this is but rarely the case. The following extract is taken from a representation made by the present Astronomer-Royal, to the Council of the Royal Society, on the advantages to be derived from the employment of two mural circles:—
"That by observing, with two instruments, the same objects at the same time, and in the same manner, we should be able to estimate how much of that occasional discordance from the mean, which attends even the most careful observations, ought to be attributed to irregularity of refraction, and how much to the imperfections of instruments."
In confirmation of this may be adduced the opinion of the late M. Delambre, which is the more important, from the statement it contains relative to the necessity of publishing all the observations which have been made.
"Mais quelque soit le parti que l'on préfère, il me semble qu'on doit tout publier. Ces irrégularités mêmes sont des faits qu'il importe de connoître. Les soins les plus attentifs n'en sau-
