MR. W. CROOKES ON THE ATOMIC WEIGHT OF THALLIUM. 
297 
boiled for half an hour. The supernatant solution is then transferred to a German green 
glass retort, and distilled over a water-bath at the rate of about one drop a minute. 
The end of the neck of the retort is drawn out, and fitted tightly into the mouth of the 
flask used as a receiver. The first portions of water which distil over are always to be 
rejected, and the distillation is stopped when about three fourths of the contents of the 
retort have come over. The product of this distillation is then mixed with freshly pre- 
cipitated oxide of silver, and allowed to stand, with frequent agitation, until a few drops 
of the decanted clear solution exhibit an alkaline reaction. The water is then introduced 
into an apparatus shown at fig. 1, Plate XVI. a, b, c are three globes made of hard 
German glass, a and c holding about half a pint each, b holding about four ounces ; 
they are connected by tubes as shown, the tubes being contracted at the places d , e,f. 
The water, containing a little oxide of silver dissolved in it, is introduced through d into 
the globe a, until a is nearly full. The contracted part at d is then sealed up before 
the blowpipe, the end g is put into connexion with the Sprengel pump from which the 
U-tube containing sulphuric acid has previously been removed, and the mercury set 
running. Heat is applied to the globe a containing the water, until gentle ebullition 
(under diminished pressure) sets in, the globes b and c being kept cold with ice. This 
is continued until no more air, either from the apparatus or from the water, is carried 
down by the falling mercury. When this is the case, the globes is allowed to cool, and 
the tube is sealed up at f, the vacuum being maintained as perfect as possible. By 
inclining the tube and globes, any water which may have distilled over from a into b or 
c , is now poured back into a. The two end globes a and c are then placed in water- 
baths, whilst the centre globe, b , is immersed in melting ice. A gentle heat being 
applied to the water-bath containing a, distillation rapidly proceeds without actual 
ebullition, and water condenses in b , its condensation in c being prevented by warming 
the water-bath in which c is immersed. When about one fourth of the contents of a has 
thus distilled over, the operation is stopped, and the bulb b removed by sealing the con- 
tracted part of the tube at li ; any trace of ammonia which might happen to be in the 
water as introduced into a will thus have been collected in the bulb b and removed. 
The apparatus now has the appearance shown at fig. 2, Plate XVI. The globe c is now 
cooled in melting ice, and gentle heat being applied to the bath containing the globe a, 
distillation again proceeds, the condensation this time being into c. The first portions 
of water which come over into c are used to rinse out that globe, a dexterous movement 
throwing it all over the inner surface without throwing any of the liquid out of a. After 
two such rinsings, distillation is allowed to proceed without ebullition until four fifths of 
the water has distilled from a to c. The tube is then sealed at the contracted part e, and 
the globe c (fig. 3), containing what I believe to be absolutely pure water in vacuo, may 
be set aside for future use. It will be observed that the water, almost chemically pure 
to begin with, has in this manner been distilled and further purified in the entire 
absence of atmospheric air. When some of this water is required for use, the glass 
tube is touched at f with a blowpipe-flame. As soon as the glass softens, the atmo- 
