261 
1914-15.] The Eeflexes of Autotomy in Decapods. 
ment causes the realising of functional capabilities which can have been of 
no practical value for so long ? 
Workers in experimental morphology (Entwickelungsmechanik) contend 
that the fundamental functional characteristics of cells are practically the 
result of “ hereditary inertia.” Arguing on this basis, therefore, we may 
conclude that the receiving cell in the unisegmental reflex of autotomy, 
which normally receives and transmits its impulse to a neuron (efferent) of 
the same level, is short-circuiting a “ current ” which in the earlier history 
of the race passed upwards to jaws and chelipeds. The resistance to 
passage of the impulse must be increased by disorganisation of the lower 
arc when the reaction-type is changed, as in removal of the hermit from its 
shell, and the impulse will then travel to higher arcs. On the nature of 
this increased resistance it would not be safe to speculate at present. 
Sherrington and others, describing the reflexes of the nervous system, 
emphasise the purposive nature of the reactions. How, then, are we to 
regard the reflex abandonment of limbs by these decapods ? In lower 
forms, evasion of an enemy seems to be the chief end, for movements of 
flight accompany local changes in the limb muscles. This is the case in the 
Natan tia and in the Palinura. It also occurs in the Anomura. Another 
element is seen, however, even in the prawn, for the next sound joint to the 
cut surface is always sharply flexed or extended, and haemostasis is thus 
produced (flg. I). Prevention of haemorrhage is also a potent factor in 
producing autotomy in the Palinura, for at the ‘‘ seat of election ” the 
animal has better venous valves to prevent loss of blood than at other 
parts of the limb. In the Brachyura the valves are most highly developed, 
and, as the causes of loss of limbs in Garcinus were shown to be the stony 
character of the shore and winter storms, so it may be concluded that auto- 
tomy normally takes place to stop bleeding. Psychic ” and ‘‘ exuviate ” 
autotomy have been described by other writers, but the evidence here 
adduced from many thousands of experiments is that evasion and haemo- 
stasis are the fundamental ends served by the self-amputation of limbs in 
decapods. Autotomy, too, is always the outcome of a nocuous stimulus 
applied to the leg which is abandoned, and it can never be proved to happen 
in the absence of such. 
I have pleasure in expressing indebtedness to Professor D. Noel Paton 
for his continued sympathy and encouragement in the work, and for his 
sound advice on such parts of it as were carried out in the Physiological 
Department of Glasgow University. 
Mr K. Elmhirst, superintendent of the Marine Biological Station, 
