98 STUDIES IN GENERAL PHYSIOLOGY 



curvature. The wall on the outer concave side of the tube is 

 therefore permanently shortened. It might seem that the 

 limit of elasticity of the tube is so low that it retains, like a 

 piece of lead, a curvature imparted to it through the muscular 

 force of the animal. But this is not the case. I put a thick rod 

 of lead into the straight tube of a Spirographis and bent it till 

 the tube was strongly curved. The lead rod was allowed to 

 remain in the tube. When a week afterward I withdrew the 

 rod from the tube, it retained only a trace of the curve 

 impressed upon it. Similar failure followed my attempt to 

 straighten by the same method, a heliotropically curved 

 tube. Yet, as I have already shown, Spirographis is able to 

 straighten its curved tube within a few hours after a change 

 in the direction of the rays of light, and, what is more, the 

 tube remains straight. The tube retains its curvature even 

 after it has been split open. The animal has, however, 

 besides pressure and pull, another means at its disposal to 

 change permanently the orientation of the tube, namely, the 

 production of a secretion and the formation of a new layer 

 within the tube. The idea that permanence in the curvature 

 is attained in this way is supported by the fact that the 

 inner layer of the tube is much more elastic than the outer 

 layers, so that the formation of a new inner layer on one 

 side of the tube might curve it permanently. The following 

 fact supports this view : If a tube is cut open lengthwise, 

 the cut margins roll inward. If the individual layers are 

 separated, as can be done easily, it is seen that the tendency 

 of the inner layers to curl up is greater than that of the 

 outer layers, and that of the innermost, newest layer is the 

 greatest of all. The formation of a new inner layer on one 

 side of the tube would, therefore, be sufficient to maintain 

 the curvature of the tube permanently. The formation of a 

 new layer cannot be observed directly. One is also disap- 

 pointed in the hope of finding one side of the wall of the 



