142 LIBBIE H. HYMAN 



pi'ostomiuiii, aprostomic condition ('rounded ends' of Krecker), 

 and finally the acephalic condition, where no tissue grows out at 

 all. All pieces from which more than seven anterior segments are 

 removed remain acephalic. 



As regards posterior regeneration, all levels of the body re- 

 generate normal posterior ends, with the usual exception that 

 the head itself remains tailless unless it includes a certain mini- 

 mum number of body segments. 



6. Sum mar ij 



It thus appears from these data that the oligochaetes investi- 

 gated form a series with regard to head formation. In Dero, 

 normal heads are formed at all levels of the body, in both short 

 and long pieces; in Lumbriculus, normal heads occur at all levels 

 in long pieces, but in short pieces, head formation exhibits pro- 

 gressive inhibition along the axis; and in the tubificids, head 

 formation is limited to extreme anterior levels, regardless of size 

 of piece. It seems to me that the earthworm and its relatives 

 occupy an intermediate position between Tubifex and Lumbric- 

 ulus. Thus in Allolobophora foetida described by Morgan ('97j, 

 regeneration of the head occurs with increasing difficulty back to 

 the level of the twelfth to fifteenth segments; behind this level 

 is a region where pieces either remain headless or produce bi- 

 axial tails, and posterior to this all pieces remain headless. In 

 the tubificids, head formation ceases at anterior levels without 

 any intermediate production of biaxial tails. 



The regeneration of the anterior end exhibits, then, these 

 striking axial differences which are of a qualitative nature. On 

 the other hand, the regeneration of the posterior end occurs at 

 all levels of the body in all of these oligochaetes, and with all 

 sizes of pieces (with the exception of the head piece, which re- 

 mains tailless unless of a certain size). The axial differences in 

 posterior regeneration are quantitative merely. 



