Experiments in Grafting 295 
In grafting or in splitting individuals, tentacles are sometimes 
displaced from their normal position, and may come to lie at 
some distance from the oral end. This result may easily be 
obtained by cutting off the head just below the line of the ten- 
tacles. When the new trunk is formed individual tentacles often 
become drawn out of place (Fig. 19). In all these cases the 
tentacles become absorbed after a time. The result is impor- 
tant, because the special conditions in the case of hydra preclude, 
with much probability, the explanation of the absorption as due 
to food relations. The tentacles are hollow and open directly 
into the central digestive space. A tentacle attached at the side 
is situated in as favorable a place for receiving food as are those 
of the ring, yet it is absorbed. The result is similar to the 
absorption of small grafts in the same position, and the expla- 
nation of the absorption is probably the same in both cases. 
One of the most important results obtained by grafting hydra 
is the reversal of polarity that can be brought about in the fol- 
lowing way: If two pieces are united by their anterior or oral 
ends, and then one of them is cut off near the line of union 
(Fig. 20), it will often produce, if very short, a mouth and ten- 
tacles (head) from its exposed aboral or posterior end (Fig. 21). 
Thus instead of a foot a head develops. This result occurs only 
when the piece is very short. It seems that the polarity has been 
reversed, owing to the union with a larger piece. The result is 
difficult to explain; possibly the conditions of tension may actu- 
ally be changed in the smaller piece through the influence of 
the larger piece, so that the orientation is reversed. The ex- 
periment needs very careful reéxamination before we can safely 
offer a probable explanation of the result. 
The planarians are also capable of remodeling the proportions 
of the old parts to produce a new worm of typical proportions. 
In this case, however, the new head and tail are formed in pro- 
liferated material, but the rest of the body is remodeled. Ab- 
normally placed parts, resulting from grafting or from cutting, 
do not show the power of pinching apart exhibited by hydra, but 
may after a time be absorbed, especially if small and not sup- 
