MEMOIR ON EMERY. 



29 



tion, first pointed out by M. de Senarmont, and which consists 

 in the isomorphism of diaspore and the hydrated oxide of iron. 

 Thus, while the peroxide of iron or oligist iron is isomorphous 

 with alumina or the corundum, the hydrates of the same oxides 

 are isomorphous. 





OQ 2. -n 



£ » o 



i" Hj "1 



: &.fD 



wff2 



&] <V ™ 



S'3^ 



0Q S O 



': 1 8, 



1 rt> = 5° 



'** ^ to 9 



': m -" J i 



Hydrated Ox- 

 ide of Iron of 

 Cornwall — 

 Dufrenoy 



M: M : 



130° 



129° 54 / 



130° 2' 

 125° 17 / 

 125° 18' 

 114° 58' 

 104° 

 151° 35' 

 151° 33' 

 116° 18' 

 167° 6' 



130° 57' 



M: bi 













M: el 



115° 

 104° 12' 



151° 36 / 







x & 



bi : e 1 







bl : bl 



151° 36' 







151° 34' 





116° 38' 







bl : c3 









144° 40' 

 145° 40' 

 126° 12' 

 116° 56 / 

 117° 46' 







e2 . p-2 















126° 20' 



p-1 : i 







116° 55' 



e^ : e^ 







117° 10' 



I have found crystals of diaspore in hydrated oxide of iron, 

 the needles traversing the oxide in all directions. There is a 

 specimen in my possession composed of a small crystal of dias- 

 pore, surrounded by a kind of scabbard of crystallized gdthite. 

 One of the summits of the crystal is exposed. In breaking 

 the oxide of iron which contains these crystals they become 

 detached, leaving on the oxide an impression with a very bril- 

 liant surface. 



The diaspore of Gumuch-dagh is also found, of a lamellar 

 structure, but very rarely; that of ISTaxos, Nicaria, and Samos 

 are all lamellar. Yet of all the specimens that I have collected 

 none offer so much interest as those composed of diaspore im- 

 bedded in corundum. Here we see the two minerals passing 

 one into the other, without being able in many places to dis- 

 tinguish the line of separation, so imperceptible is the grada- 

 tion. After what has been said in respect to corundum it is 



* The values of these angles are just as given by the goniometer, without 

 any correction by calculation. 



