194 
PROCEEDINGS OF THE NATURAL [APRIL, '89. 
reduction with Na it is claimed that alumina would form 
. upon the surface of the minute globules of aluminium and 
thus prevent cohesion between them. Cryolitewas usedas 
a flux to assist in removing this oxide. Prof. H. Rose first 
used cryolite (the double fluoride of sodium and aluminium 
Al 2 F 6 e NaF) as a source of Ai. He Heated cryolite 
with Na m an iron crucible, using KU as a flux, and 
stirring with an iron rod. His metal of course always 
contained some iron. Rose never obtained more than 80 
per eem. of the theoretical amount of A1. in the cryolite, 
and many times not more than 33 per cent 
Eassetaffirn’ecl that all metals will decompose AleClo 
morfi , ao " Wch form 'Mori*, that are 
arsenic bo"'" ° r Ta P° ri *«i these salts. Hence 
icl a,,t ' mony < S '» <Hg) and Zinc (Zn) 
be the best 7 “T age,,ts °> Al Zinc would seen, to 
aratino fr " lts cheapness and the greater ease of sep- 
t tS all0 ' V ° f AI b >- distillation or the zinc. 
Parts of Ain Cl.“' ^“3°” COnsist8 . i " 3 
r ass gets de n i^r:„li^ 
"°“ r ° r ”T 
salt poured off after the solidif ■ 5 &n ' r ° n r ° d a " d 
is obtained eonsis Lg of e ' u '!f° f l , h ' ”«*'• Aa ^ 
Basset’s process w-,/,- q et l uiv alents of Al and Zn. 
ers. Specht in I860 ^ C ° mmended V Wedding and oth- 
Za a„d P rec„mme“ed it rsZrl C ' a * *“ - af 
Dnllo states that there is ‘JTV 8 P rartl “ 
double chloride with 7n W > - dlfhcult y in reducing the 
and the loss of Zn 1 ls not a s easy as with Na, 
turc squired also burns a great ^ ^ tempera ' 
penence is that the reduction is ver" ^ AL My ^ 
a very small percentage of the lncom P lete and onl - v 
can be obtained t hr v * c °ntained in the salt 
The I«x*« has been more recently 
