320 Chittenden and W hitehouse—Metallic 
Srriss III. 
la 0:5860 0:0239 4:07 
b 0:6949 0:0284 4:08 
2a 0:7090 0:0290 4°09 
b 0:90538 0:0374 4°13 
3a 0:5980 0:0247 4:13 
b 0:8000 ()0328 4°10 
Series IV. 
la 06217 00300 - 4°88 
b 06810 00331 4:86 
2a 05509 0:0270 4-90 
b 05626 0-0278 4°86 
3a 0°6955 0°0396 5°69 
b 0°7515 00430 9°72 
The figures show a far smaller content of silver in all of the prep- 
arations than found by Lieberkiihn or Mulder. In three of the 
series, there is seen a constancy of composition which is quite notice- 
able and, further, a close agreement with the second result obtained 
by Loew. on adding a 5 per cent. solution of silver nitrate to the 
albumin. In the last series, however, the percentage of silver is 
somewhat higher, possibly owing to incomplete dialysis of the 
chlorides and phosphates from the albumin solution. These figures, 
however, are not much higher than the highest figures obtained by 
Loew. 
A silver salt of albumin, of the composition (C,,H,,,N,,SO.,), 
+ Ag,—H, would contain 4°28 per cent. of Ag, and while our results 
certainly approximate to this figure, there is variation enough to 
indicate an equal possibility of a mixture of two or more com- 
pounds. With a molecule of the size of the albumin molecule, it is 
possible by doubling or otherwise, to obtain a formula corresponding 
to almost any percentage of metal found. And inasmuch as every 
variation* in the method of preparing the albuminate tends to alter 
its composition, it seems worse than useless at present, to lay much 
stress on the exact constitution of the silver albuminate. A large 
number of albuminates are of course possible, but until we know 
more definitely how to separate one from another, we have no guaran- 
tee of the simple nature of any one. 
* Loew states that he has prepared a silver albuminate containing 10°7 per cent. 
Ag, corresponding nearly to 6 atoms of silver, and that it is possible to prepare album- 
inates still richer in silver. 
