312 
MR. W. CROOKES ON THE ATOMIC WEIGHT OP THALLIUM. 
into the apparatus where the subsequent operations are to take place. The pieces of 
tube, with any splinters which may have broken off, are weighed, first in air and then in 
a highly rarefied atmosphere. The difference between the weighings of the full and 
empty tube, after correcting for the hydrogen contained at first, gives the weight of 
thallium taken. 
These two forms of apparatus were found to answer the purpose tolerably well. 
Several improvements, however, suggested themselves whilst the determinations were in 
progress, and they were finally embodied in the apparatus shown in Plate XVI. figs. 
11, 12, & 13. In this several refinements of manipulation can be introduced which were 
impracticable with the former apparatus — notably the ease with which a definite quan- 
tity of metal is introduced into the apparatus without the chance of oxidation, the sim- 
plifications introduced in the weighings consequent on having the apparatus vacuous, 
and the facilities obtained for the employment of the Sprengel and Bunsen pump at 
different stages of the operations. 
Although each determination with this improved apparatus still took many weeks for 
its successful performance, a great saving of time was effected when compared with that 
required for a determination in the apparatus first used, where months were consumed 
in the evaporations. As this form of apparatus was the one in which most of the deter- 
minations were effected, and as the manipulations were attended with greater chances 
of accuracy than Avere those at first employed, I will describe the apparatus, its employ- 
ment, and the several processes performed in it somewhat in detail. 
Some of the metallic thallium prepared by one of the methods already described is 
cut by means of a sharp steel knife into prisms about one eighth inch square and half 
an inch long, no particular care being taken to avoid oxidation. The prisms are boiled 
in dilute hydrochloric acid to remove any trace of iron which the knife might have 
communicated. They are then washed in w T ater, dried with blotting-paper, and intro- 
duced into the cylindrical portion «, fig. 11, of the apparatus. The outer extremity of a 
is then draAvn out and sealed before the blowpipe. The end c is also sealed up and the 
horizontal tube e is connected to the Sprengel pump and a vacuum obtained, the appa- 
ratus and the thallium being kept warm to drive off any moisture which might have 
been introduced with the thallium. When the vacuum is perfect the tube is sealed 
at f. The apparatus, sealed up and entirely free from air, is noAV laid on its side, and 
the cylinder a and the bulb b imbedded in a bath of magnesia held in a copper vessel 
heated by gas. The temperature is then raised to above the fusing-point of thallium 
(561° F.), when by careful manipulation the oxide may be separated from the liquid 
metal and the greater part of the oxide collected at the closed end of the cylinder a. 
The magnesia is then removed from about the narrow part of the tube cl (which should 
be somewhat long and very much contracted), and by a dexterous movement the mag- 
nesia-bath containing the apparatus is suddenly tilted up and the liquid metal allowed 
to run through the contracted part into the bulb b. In some instances portions of oxide 
or of metal stick in the channel, then the operation is lost and a fresh attempt has to 
