METHOD FOR DETE;RMINING WEIGHT OE PARTS OF EGGS. Ill 



directly opposite to what would be expected from the slight 

 difference in weight in the two groups. The mean weight of 

 the eggs separated fresh is slightly less than the mean of the 

 other group. It would therefore be expected that the mean 

 percentage of albumen would be less and the percentage of 

 yolk greater than in the other group. This difference is prob- 

 ably due to chance individual variations although it is possible 

 that the yolk actually gains weight at the expense of the albu- 

 men when the egg is boiled. 



The mean percentage of shell before correction, and even 

 more after, is slightly less in the boiled than in the fresh eggs. 

 This is what we should expect if the separation between this and 

 the albumen were incomplete. This is, indeed, the most prob- 

 able explanation of the fact. The fresh shell has a damp 

 appearance even after it has been wiped as dry as possible with 

 filter paper. The boiled shell looks almost dry. It is possible 

 that the shell is somewhat dissolved in boiling and also possible 

 that some of the difference is due to chance individual variation. 

 All the work on the weights of the parts of eggs done at this 

 laboratory shows that the chance individual variation in shell 

 weight of eggs of equal weight is very large. 



To summarize this portion of the work it may be said that 

 a comparison of the method of separating the parts of eggs 

 fresh with the method of separating them after boiling shows 

 that the error of weighing (i. e., the difference between the sum 

 of the weights of the parts and the weight of the whole) is not 

 essentially different in either case. Certainly this error is not 

 larger when the eggs are separated fresh. Further, when the 

 eggs are separated fresh there is no appreciable loss of albu- 

 men due to drying. When eggs are separated after boiling 

 there is in addition to the irreducible error in weighing an 

 error five times as large, viz., the loss in weight due to the boil- 

 ing and cooling. This error is not distributed pro rata between 

 the different parts of the egg but is largely a loss to the albu- 

 men. The data here studied indicate that the yolk may even 

 gain weight at the expense of the albumen although the range 

 of variation and the small number of eggs considered make a 

 positive conclusion on this point impossible. 



The weight of the shell can undoubtedly be found a Iftfle 

 more accurately in the boiled egg. Even when the shell of a 



