CAUDAL AUTOTOMY AND REGENERATION IN THE GECKO 7] 
ment of the regenerated tail m H.flaviviridis up to that 
of seven weeks, and Text-fig. 5, J’, K’, L’, shows a second tail 
bemg regenerated on the broken-off stump of a first regenerated 
tail. 
The exact length of time it takes nm H. flaviviridis 
for a new tail regenerated from the base to attain the full length 
of the original tail I do not know, but it is certainly not less 
than four or five months, and is probably more. 
(7) There is apparently no limit (save that imposed by the 
longevity of the animal) to the number of times a tail can be 
regenerated. 
(8) The skin of the regenerated tail is not a mere extension 
of that of the original tail but is a new product, as shown by 
both lepidosis (‘Text-fig. 1, D) and texture. The skin of the 
original tail is, like that of the trunk, head, and limbs, very soft 
and rubs off easily (the tail in consequence not being easy to 
skin), whereas the skin of the full-grown regenerated tail is 
relatively tough and the tail is easily skinned. After auto- 
tomy the original skin shows no signs whatever of growing 
over the raw exposed surface, and remains quite distinct from 
the new skin which covers the outgrowing regenerated tail 
(Text-fig. 1, H—P). 
Tur GENERAL STRUCTURE OF THE ORIGINAL TAIL OF 
Hemidactylus flaviviridis. 
In those Lacertilia in which the tail is of distinct use to the 
animal for purposes other than that associated with autotomy, 
e.g. for prehension (as in the Gecko Ceratolophus auri- 
culatus, Bavay, and in Chameleons), for swimming (as in 
aquatic Monitors, Iguanidae, Amblyrhynchus, Lophurus, and 
Physignathus), for steering (Basiliscus in water, Ptychozoon 
in air), or for balancing in air (Draco), caudal autotomy naturally 
does not occur, but it appears to me that the tail in the more 
common Lacertilia (Lacertidae, Agamidae, &¢.) can be of but 
very little use to its owners. It is in these forms not used for 
swimming (as may be proved by throwing lizards into water, 
