362 G. E. GATES. 



extensively while anterior portions longer than one-half never 

 autotomized. Several posterior portions about equal to one-half 

 the length of the original worm autotomized from each end one 

 or two pieces which died while the longer middle portion survived. 



The experiments were discontinued before a thorough analysis 

 of various aspects of this interesting tendency to fragment could 

 be completed. Two experiments, however, provided a hint, not 

 only as to the cause of the fragmentation but also as to the cause 

 of certain other disturbing factors. Worm 94 was anaesthetized 

 as usual and cut into three approximately equal portions, each of 

 which was kept separately in a tightly closed jar. No autotomy 

 occurred in any of the jars. The head piece survived and regen- 

 erated a tail. The middle piece regenerated at one end an anal 

 segment and at the other end a head about three millimeters in 

 length composed of fifteen segments. On the day following the 

 operation the tail portion was collapsed and flattened, dead, but 

 without noticeable odor of decay. In the jar were three flies 

 which must have been present in the tail portion of the worm at 

 the time of the operation, as the jar was not opened until after the 

 flies had appeared. Through the kindness of entomologists at the 

 Imperial Bureau of Entomology, London, these flies have been 

 identified as Aphiochata scalaris Lw. 



A number of head portions ranging from twenty-five segments 

 to about half the length of the worm were kept in a single, large, 

 tightly-closed jar. A few portions died during the first three days 

 after the operations and were removed. At the end of the week 

 four of the head pieces were still living and apparently in good 

 health although without signs of regeneration. The only traces to 

 be found of the other head portions were numerous tubular frag- 

 ments of transparent cuticle. Crawling around inside the jar 

 were numerous small insect larvas. When the jar was opened 

 two small flies very similar in appearance to those from the other 

 worm escaped. There seems to be no reason for doubting that 

 some larval stage of the fly was parasitic in these worms at the 

 time of the operation and it is at least possible that the presence 

 of parasitic fly larvae is the factor responsible for the autotomy as 

 well as other disturbances in the operated worms. It should be 

 noted, however, that A. scalaris is a very general feeder and that 



