use of the rejecting goniometer; these agree in the incidence of P on P as 136° 47′, given by Haüy as 137° 10′. These crystals are very small, and as they differ from each other, and exhibit the planes of some modifications not hitherto described, I have not found courage to run the hazard of sacrificing them, incidental to the attempt to cleave them in the direction of their natural joints.
Specular Iron.
This substance, a variety of the fer oligiste of Haüy, may be split in the direction of its natural joints when held in the hand, by means of the pincers, care being taken to place their edges parallel with the primitive planes, which are generally observable on the edges of the crystals from Elba. In this manner I have procured six fragments, one of them the primitive rhomboid, which is slightly acute, nearly complete, and all of them having two or more planes sufficiently brilliant for the use of the reflecting goniometer, and all afford the measurements of 86° 10′ or 93° 50′, some of them both; the former being 59′ less, the latter 59′ more than the measurements obtained by Haüy on the natural planes by means of the common goniometer. The perfect agreements afforded by the fragments, have so far satisfied me that the results are the true value of the angles of the primitive rhomboid, that I have not attempted to measure its angles by means of the reflections to be observed on the natural planes, which are often very brilliant; for experience has confirmed me in the opinion that as the natural planes do not often yield results agreeing amongst themselves, they cannot be relied on with the same confidence as the planes obtained
