6 C. M. CHILD. 



wall, thus producing a condition in which the oesophagus opens 

 aborally as well as orally to the exterior and the enteron is com- 

 pletely closed. As will be shown at another time, regeneration 

 in such pieces is always slight. 



The smallest pieces in which normal regeneration has been ob- 

 served were about one twentieth of the body-length and from 

 the region just posterior to the oesophagus. When the regenera- 

 tion was complete these pieces appeared to consist of little but 

 tentacles and disc, which were only slightly smaller than in 

 pieces many times as large. My observations prove that typical 

 regeneration is possible in very small portions of the body, pro- 

 vided they represent complete transverse or oblique sections of 

 the body ; pieces in which the transverse continuity is inter- 

 rupted roll up in such manner that typical regeneration can 

 never occur. 



The Effect of Temperature on Regeneration. 



No special experiments in relation to temperature were per- 

 formed, but my work upon Cerianthus extended from September, 

 1902, to February, 1903. In September the temperature of the 

 water was very high ; during October, as the weather became 

 cooler, it fell gradually and continued to do so as long as ob- 

 served, i. e.,, until February. A comparison of similar series of 

 pieces begun at different times during this period affords an inter- 

 esting illustration of the effect of temperature. The various 

 series for which data are given in this paper will serve for this 

 purpose.] 



Series 22 was begun September 24; Series 35, October 20; 

 Series 45, November 7 ; Series 54, 55 and 56, December 15. 

 If the series be considered in this order a considerable decrease in 

 the rapidity of regeneration is noticeable. In Series 22, Septem- 

 ber 24, the marginal tentacles first appeared in the pieces A four 

 days after section. In Series 54, 55, 56, December 15, the mar- 

 ginal tentacles in the pieces A from about the same level of the 

 body as Series 22A appeared eleven days after section. Between 

 these two extremes lie the other series, and in every case the 

 later in the season the series was begun the less rapid the regen- 

 eration. It is not necessary to go over the data in detail to 



