94 



PROBLEMS OF RELATIVE GROWTH 



I have (last column) calculated the ratios of ? to £. 

 These indicate clearly that both for length and breadth there 

 exists a growth-gradient in the ? abdomen, with high point 

 in the sixth segment, though the two gradients must be of 



very different 

 slope and shape. 

 Precisely simi- 

 lar results have 

 been obtained for 

 the female ab- 

 domen of the 

 common Carci- 

 nus maenas (un- 

 published). That 

 the growth- 

 centre need not 

 be in the pen- 

 ultimate seg- 

 ment, however, is 

 shown by the 

 Pea-crab, Pin- 

 notheres pisum. 

 Scale drawings of 

 this at different 

 stages of its 

 growth are given 

 by Atkins (1926) ; 

 I have measured 

 these, and al- 

 though the num- 

 ber of specimens 

 is very small, it 

 is quite enough 

 to demonstrate 

 the gradient. The 

 gradient is first 

 of all interesting 

 because of its steepness. The growth-coefficients of the separ- 

 ate segments, relative to carapace length, range up to 2-3 

 and over, a very high figure. In passing, this remarkable 

 heterogony is doubtless correlated with the small absolute size 

 to which the crab is restricted within its host's shell ; the 

 animal has to attain its full female proportions at a much 



Fig. 53. — Changes of shape in the female abdomen of 

 the pea-crab, Pinnotheres, during growth. 



Above, stage I, late (carapace width about 3 mm.) ; centre, stage 

 II 16 (carapace width 6 mm.) ; below, final form, stage V (carapace 

 width 1 1 -5 mm.). Thp abdomen is at first male-type, but shows 

 marked heterogony, with terminal growth-centre, until it overtakes 

 the bases of the legs. 



