12 J. WILLIAM BUCHANAN 



The remaining worms were cut in the same manner in groups 

 of fifty, but the pieces were dropped as soon as cut into the appro- 

 priate concentration of the anesthetic. As soon as the cutting 

 was completed, the solution was replaced by fresh and the flasks 

 stoppered as the control. One series of A, B, and C pieces re- 

 generating in water thus served as a control for one or more 

 series subjected to the anesthetic. In experiments designed to 

 test the effect on head frequency of short-period exposures to 

 the anesthetics the solutions were poured off after a number of 

 hours, the pieces washed several times and set aside to regenerate 

 in water. In experiments designed to test the effect on head 

 frequency of more dilute solutions of the anesthetics applied 

 throughout the period of regeneration, the anesthetic was not 

 replaced by water. In all experiments the water or solutions 

 were replaced by fresh every forty-eight hours. Both control 

 and experimental series were kept on the same shelf at approxi- 

 mately the same temperature and out of direct light. At the 

 end of fourteen days the regenerated animals were removed from 

 the flasks and examined under a dissecting microscope. The 

 number of each type of anterior end was tabulated in per cent, 

 and the result for each flask is considered the head frequency 

 under the conditions of the experiment of the level from which 

 the pieces were cut. 



Slight differences in length of the pieces undoubtedly intro- 

 duce an unavoidable error. Another source of error lies in slight 

 differences in the physiological conditions of the animals. The 

 total error is certainly not greater than 10 per cent, but a dif- 

 ference between the control and experimental pieces from the 

 same level of less than 10 per cent that repeats itself throughout 

 a series of comparable experiments must be considered as evi- 

 dence of the effect of the anesthetic on head frequency. To 

 reduce the subjective error of cutting pieces of unequal length 

 as much as possible, in some of the earlier experiments the con- 

 trol and experimental sets were cut in alternate groups of ten. 

 However, no differences in results could be noted between such 

 series and those cut in two consecutive groups of fifty and the 

 method was discontinued as time consuming. 



