190 Chittenden and Lambert— Post-mortem Formation 
beyond the ordinary limits of error. This question we have en- 
deavored to answer in the next experiment. 
Experiment IV. : 
A rabbit’s liver removed from the body immediately after death, 
was prepared in the usual, manner. Two mixtures, exactly alike, 
were then made as follows: 
A. B. 
25 grams of liver. ~ 25 grams of liver. 
100 ec. c. of water. 100 c.c. of water. 
These were in the bath 23 minutes after the death of the animal — 
and were warmed at 40° C. for 2 hours. The two portions were 
then extracted and analyzed as in the preceding experiments; the 
object being to see how great a variation would be obtained by this 
like treatment of the two portions of sampled liver. Following are — 
the results: 
Glycogen A. 
Volume Equivalent Equivalent Total Per 
used. Weight Cu. in dextrose. in glycogen. amount. cent. 
25 ©. ¢. 0°1575 gram. 0-0802 gram. 0°0721 gram. 0°5768 gram. 5°76 : 
Glycogen B. 
25 ¢, ¢C. 0°1585 gram. 0°0809 gram. 0°0728 gram. 0°5824 gram. 5°82 : 
Sugar A. | 
10 ¢. ¢. 0°0433 gram. 0:0225 gram. == == 0°2250 gram. 2°25 
Sugar B. : 
WO) 5 0°0430 gram. 0:02240orant, |=) (she. 0°2240 gram. 2°24. 
Total carbohydrates A. 
25 ¢..C. 0:2340 gram. Oa ZOReetams Siete 0°9656 gram. 9°65 
Total carbohydrates B. 
25 ¢. ¢. (2335 gram. Op1203 jorams: se | esses 0°9624 gram. 9°62 
Percentage results. 
Glycogen. Sugar. Total carbohydrates. 
A. 5°76 per cent. 2°25 per cent. 9°65 per cent. 
B. 5°82 2°24 9°62 
—0°06 +0°01 + 0°03 
These results plainly show that when the conditions of the exper- 
iment are exactly the same, the average variation in results will be 
considerably less than 0°1 per cent. Consequently variations greater 
than this must have their origin in something other than the ordinary 
errors of analysis. Hence, in the three preceding experiments we 
have to account foran average increase of about 0°5 per cent. in 
