D.— ZOOLOGY. 63 



smaller contraction when the new molecule is formed, so these fragments 

 of protein pulsate steadily through the day. So they continue through 

 the ages, while protein enters into new combinations, and the aggregate 

 of protein molecules is replaced by a unit of protoplasm, still keeping the 

 rhythm of saving up energy and making-a-molecule, saving up energy and 

 making-a-molecule. 



Now protoplasm in most organisms which we can study becomes 

 altered at the surface which is in contact with water, by a change which is 

 conveniently called ' gelation,' the protoplasm at the surface losing most 

 of its fluidity and changing in other properties. In certain circumstances, 

 such as increased salinity of the water, the internal fluid protoplasm will 

 burst out through this gelated surface in fine threads, which either gelate 

 in their turn or change into strings of drops. 



I venture to suggest that the great evolution of the flagellate, which 

 Church pointed out to us, accomplished in some ten thousand or hundred 

 thousand million generations, was the formation of a permanent filament 

 of protoplasm of which one side was more gelated than the other side, so 

 that one longitudinal strip of the cylindrical outer surface is more elastic 

 and therefore less easily extensible than the opposite strip. Let us sup- 

 pose the gradual accumulation of energy causing, as before, a gradual 

 increase in volume of the protoplasm ; then the more easily extensible 

 surface will swell, and therefore lengthen, and the filament will gradually 

 bend. When the quantum of energy is reached which suffices for forma- 

 tion of a new molecule, every old molecule will suddenly lose its surplus 

 energy and return to its old molecular volume, the distended surface will 

 return to its old dimensions and the filament will straighten. 



I have spent an appreciable part of my life watching the flagella on the 

 living collar-cells of Calcareous sponges— Grantia, Sycon, Lencandra, and 

 Clathrina. Their movement is nearly confined to one plane and is asym- 

 metrical, being almost always with a faster beat to one side than to the 

 other. There is a pause, a stroke and a counterstroke. Mr. James 

 Gray pointed out to me that if the counterstroke be elastic, as I supposed, 

 it should always take the same time, as compared with the varying time 

 of the active contraction. This I found to be the case. At about 

 24 double vibrations to the second, the stroke and counterstroke are of 

 equal duration ; at higher frequencies the stroke is the shorter, as in 

 a schoolmaster's cane ; at lower frequencies the stroke is the longer, as in a 

 fisherman's trout-rod. The broad features of the phenomena are therefore 

 consistent with the hypothesis that the counterstroke is an elastic 

 rebound. 14 



The. apparent improbability of a lowly organised cylindrical cell, with 

 an axial straight flagellum, having one longitudinal strip of the surface 

 of that flagellum different from all the rest of that surface, disappears 

 when we recognise that one longitudinal strip has a different history from 

 all the rest of the surface. A collar-cell in a sponge is usually surrounded 

 on all sides by six other collar-cells, of which one is its twin sister. Like 

 all flagellates, including metazoan spermatozoa, collar-cells divide longi- 

 tudinally. The details of this division were worked out very beautifully 



14 See Appendix B. 



