THE AXIAL GRADIENTS IN HYDROZOA. 387 



greater importance. In experiments 32 and 33, the longer pieces 

 in each case regenerate faster, but the effect of length does not 

 overcome the effect of level, the apical pieces in both experiments 

 regenerating first on the whole. The influence of length is most 

 marked in experiments 32 and 33, where short pieces were 

 employed. It is very little evident in experiments 37 and 38, 

 and 42 and 43, where pieces 10 mm. in length were employed. 

 In fact, in experiment 37, the apical pieces 10 mm. long regenerate 

 slightly faster than the apical pieces 20 mm. long; and in experi- 

 ments 42 and 43 there is practically no difference. We may 

 therefore say that length is of little consequence in pieces exceed- 

 ing 10 mm. in length, in agreement with Child's previous state- 

 ment and in contradiction to the claims of Banus. In all 

 experiments the longer basal pieces regenerate faster than the 

 shorter basal pieces, but it is probable that this effect is one of 

 level rather than of length, because the longer basal pieces have 

 their apical ends at a higher level than the shorter basal pieces. 



It may therefore be concluded that in the case of apical and 

 basal pieces of unequal length, the apical pieces will still regen- 

 erate more rapidly on the whole regardless of whether they 

 constitute the shorter or the longer pieces, always provided that 

 their minimum length is 5 mm. The level at which the pieces 

 are cut is still the dominant factor in such experiments. In 

 pieces below 10 mm. in size, a longer piece will regenerate slightly 

 faster than a shorter one; but in pieces above 10 mm. length, 

 length is of practically no consequence. Long basal pieces in 

 such experiments regenerate faster than shorter basal ones mainly 

 because their apical ends are at a higher level than the apical ends 

 of the shorter pieces. These results are the contrary of those of 

 Banus whose experiments are invalidated owing to his erroneous 

 method of cutting the apical pieces as discussed at greater length 

 below. 



The data already presented incontestably demonstrate that a 

 regional difference in rate of regeneration exists along the axis 

 of Tubularia, such that regeneration is the more rapid the nearer 

 the piece lies to the apical end. It may next be inquired whether 

 this regional difference is modifiable under either certain normal 

 conditions or under experimental conditions. To this inquiry 



