102 C. M. CHILD 



I. TECHNIQUE 



Since considerable numbers of individuals are used in each 

 experiment, some means of standardization of the material 

 as regards physiological condition must be found, and, as noted 

 in earlier papers, size is the best available criterion of physiologi- 

 cal condition. In each series, then, animals of the same size 

 from the same stock are used both for controls and for the 

 experimental conditions and with every experimental series a 

 control under standard conditions is made. This is rendered 

 necessary by the fact that the physiological condition of any 

 stock changes with growth, advancing age, temperature, nutri- 

 tion, etc., and since head-frequency varies with all these con- 

 ditions a control made at one time is not entirely satisfactory, 

 even for the same stock a few weeks later. 



In the experiments described below worms of large size in 

 nearly all cases 16 to 20 mm. long are used and in most cases the 

 region of the first zooid in such worms is cut into three equal 

 pieces, excluding the head. Of course with worms of different 

 size or with pieces of different size the head-frequencies will be 

 different both in controls and in experimental lots, but this 

 factor of size will be considered in a later paper. In animals 

 15 mm. or more in length the differences in length are due very 

 largely to the growth of the posterior zooids (Child, "lid) the 

 length of the anterior zooid differing only shghtly in such ani- 

 mals because it has attained approximately its maximal length, 

 while in smaller animals its length increases with, though less 

 rapidly than, the length of the animal. Except where otherwise 

 stated the comparisons are made with lots of fifty pieces each, 

 a lot including only pieces of as nearly as possible the same 

 size and from as nearly as possible the same level or region of 

 the body and parallel lots being prepared for control and experi- 

 ment. Experience has shown that this number is sufficient 

 to give fairly definite results and it is about as large a number as 

 can be handled readily. Repetition of the experiments serves 

 as a means of confirmation of the results of single series. 



