PHYSIOLOGY OF GROWTH-GRADIENTS 169 



Another alternative is to suppose that the nervous system 

 is implicated. This is suggested by unpublished observations 

 of F. N. Ratcliffe on the nervous system of male Uca. He 

 rinds a marked asymmetry in the thoracic ganglionic mass, 

 naturally largest in the ganglion of the chelar segment but 

 visible in other segments as well. There are two classes of 

 cells in the nuclei associated with the ganglia, large and small. 

 The large are usually of different sizes but the same in 

 number on the two sides of the body. But the small, at 

 least in the ventral nucleus, appear to be equally distributed 

 in number in early stages and unequally distributed (larger 

 number on the side of the large chela) in later stages. Ratcliffe 

 tentatively suggests that this condition is brought about by 

 the permanent transfer to the side of the large chela of a 

 certain number of originally median cells which could be 

 transferred to either side of the body. If this suggestion be 

 substantiated (admittedly it needs further work for its verifi- 

 cation), it would imply that the normal fixing of the male- 

 type potentiality would be finally due to a certain number 

 of ' neutral ' nerve-cells being transferred to that side, while 

 the capacity of both claws to regenerate of male type in the 

 doubly male-clawed older males would be due to a loss, with 

 age, of the capacity of the median cells to transfer themselves 

 from one side to the other. 



This suggestion would obviate the need of postulating the 

 proximal spread of a specific growth-promoting capacity in 

 the limb. It is, however, perfectly compatible with the idea 

 of an initial chemo-differentiative localization of high growth- 

 potential in the male chela. 



In this connexion also the observations of Perkins (1929) 

 are interesting. In crabs (Carcinus, Cancer) and lobsters 

 (Homarus) he finds a gradient in the body as regards the con- 

 tent of glutathione, sulphydryl and other reducing compounds 

 known to be associated with growth. And this runs parallel 

 with the actual growth-relations of the various appendages 

 and regions of the body. He further advances a biochemical 

 hypothesis to explain the existence of growth-gradients ; but 

 as this is highly speculative, and as I do not pretend to 



the other segments by repeated fission. In certain cases (1907B) he 

 was able to show that the normal directions of regenerative differen- 

 tiation could be reversed by special conditions. It is probable that 

 these facts are to be in some way related with those of growth-gradients, 

 but for the moment the connexion remains obscure. 



