714 A. Ff. Goldfarb 
Experiment 7. When a square piece, b’h, Fig. 23, including 
the ganglia and a large part of c and c’ is cut out, regeneration of 
a head takes place as we should expect. When eh is removed, 
so that e occurs posterior to X and the strip / contains neither 
ganglia nor c and c’, a head 1s regenerated, which may be either per- 
fect or imperfect. The head grows in different places depending 
on the nature of the original injury, and on the extent of the sub- 
sequent shrinkage or sloughing. If the latter takes place the 
resulting individual regenerates a head when the level is anterior 
to X, an imperfect head when close to x, and no head posterior to 
X. If the strip / is long the new head may form altogether from 
the cut side, without any apparent contribution from the posterior 
level as at d, in Fig. 22. In other words when a piece of 
sufficient bulk anterior to X, but not containing ganglia nor cords 
c and c’, is left attached to the animal, such a piece can regenerate 
anewhead. It appears to be a question not of nerve supply but of 
mass or kind of material. 
Experiment 8. ‘This experiment is interesting because it shows 
that still another factor has to be reckoned with, in anterior 
regeneration. If we remove the middle area including ganglia 
and cords, and leave two narrow bands kh andh’, Fig. 23, these 
come together, curl upon one another and may behave in any one 
of the following ways: ‘The strips may break off at their bases, b’ 
or e; if near the former level a new head is produced, if near the 
latter, no regeneration takes place. Where little or no sloughing 
occurs, lateral regeneration between the cut strips may take place 
and a new head is regenerated. Very often however particularly 
where the strips are long, 1. e., they extend back of x, the two strips 
embrace each other firmly, and no lateral regeneration occurs. 
Such animals lived for a long time but no head was regenerated. It 
was shown in a previous experiment that each of these strips has 
latent the power to regenerate laterally, yet here something pre- 
vents the piece from exercising its power. For lack of a better 
term, I have called this inhibiting influence the mechanical interfer- 
ence of parts. Stevens has found a similar phenomenon in other 
planarians. 
Experiment 9. A narrow strip 7, Fig. 23, was left projecting 
