248 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
added. The difference between the accuracy of 
measuring the fish when caught, and again when 
recaught, disappears when the data are treated 
statistically; though in the case of individual examples 
it might lead to such an anomalous result as a negative 
growth. 
(2) The fish when dead may have contracted to a 
significant extent. I discuss this question on p. 187 of 
this Report, but as a rule, in measuring recaught 
marked plaice, the fish has been ‘“‘ stretched,’’ a plan 
suggested by Mr. G. P. Farran. One holds the fish by 
the head and tail, and then pulls at it—not so strongly, 
of course, as to break it in two—and the fish slightly 
elongates. The increase in length is probably very 
nearly equivalent to the shortening produced when 
rigor mortis sets in. 
The Amount of Yearly Growth. 
If the increments in length of all the marked plaice 
recovered after one complete season’s growth, and no 
more, be added together and divided by the number of 
cases, we obtain the mean annual growth for the range 
of lengths represented. The experiments undertaken 
in the Irish Sea give 58 such cases, and the mean annual 
erowth increment is 2°836 inches, or 72 cms. The limits 
are 2°56 cms. and 13°3 cms.; the standard deviation is 

o 
: 4 5 hs rar SINGS Ee 
8°78, and the probable error, Wa — + 1:14 cms 
Cases of 10 marked plaice recovered after two 
complete seasons’ growth, and no more, are represented 
in the results, and the mean increment is 5°53 inches, or 
14°04 cms., very approximately double that of the single 
season’s growth. Only one case of a plaice recovered 
after three complete seasons’ growth is represented, and 
