28 COMPARATIVE PHYSIOLOGY 



to rest at the end of the forward stroke. If the hydrogen ion 

 concentration is increased suddenly to a much lower pH. 

 than that which suffices to bring about cessation of movement, 

 the cilia come to rest not at the end of the forward stroke but 

 in the contracted condition, i.e. at the end of the recovery stroke. 

 We see then that the cilia are brought to rest in the relaxed 

 condition at a certain degree of acidity depending on the 

 penetration of the cell by the acid ; but if a greater quantity of 

 acid is present, they are brought to rest in the contracted 

 condition. This Gray interprets as due to the fact that the 

 cell itself is more permeable to acid than the cilium. In hyper- 

 tonic solutions arrest is brought about by a reduction not of 

 the rate, as with stoppage in acid medium, but of the ampHtude 

 of the beat ; and the fact that the amplitude is affected by an 

 increase in the osmotic pressure of the external medium 

 suggests that withdrawal of water from the cell interferes not 

 with the periodic liberation of energy but with some part of 

 the contractile mechanism. 



To bring all these phenomena within the scope of a 

 single hypothesis Gray has suggested that the cilium flies 

 forward owing to imbibition resulting from periodic libera- 

 tion of acid. This is not essentially very different from a 

 suggestion put forward many years earlier by Schafer. But 

 direct proof of the production of acid in ciliary movement is 

 lacking. 



Turning now to the second aspect of the problem, that is, 

 the relation of oxygen to the contractile mechanism, there is 

 now satisfactory evidence of a close analogy between the ciliary 

 and muscular tissues. The mechanical activity of cilia can 

 be treated quantitatively by timing across a standard distance 

 of gill filament or other ciliated epithelium the movement of a 

 minute circular plate of platinum. The oxygen consumption 

 of ciliated epithelium can be conveniently measured by the 

 Barcroft manometer. When the rate of oxygen consumption 

 and mechanical activity are plotted for various temperatures 

 and reduced to the same scale of ordinates, the curves 

 correspond closely (Fig. lo), showing that the rate at which 

 oxygen is consumed is normally a function of the mechanical 



