The dates when ecdysis occurred indicated no relation with either age or sex, and the final ecdyses at which the adult condition commenced were distributed over the whole period in the same ratio as those of still immature individuals.

The statement by Cornelius, that the post-embryonic development of S. orientalis extends over several years, received distinct confirmation from my observations, though to a very imperfect extent, as the purpose for which the immature individuals were kept in captivity was not primarily connected with their lifehistory, and consequently they were usually removed and placed in spirit when they had completed a single ecdysis. With regard to the statement that an ecdysis is performed immediately on leaving the egg-case, I failed to observe the actual emergence of the young therefrom, though in several cases I examined them within twenty-four hours of their hatching out. With these

I could find no trace of cast cuticles, nor did any ecdysis occur for a week succeeding the examination. It is quite possible, however, that ecdysis did occur almost simultaneously with hatching, and that the cuticles shed were speedily devoured, as the young of this species commence to feed within a few hours of hatching. As already mentioned, the four-jointed form of tarsus was found in young only 4 cm. in body-length, which appears, from measurements I have made, to be the length of newly hatched individuals. Whether or not such very young individuals could have lived long enough to suffer loss of a tarsus, perform ecdysis, and reproduce the lost part it is not possible to decide.

As regards the statement that two further ecdyses are performed in the first year, I was able only to ascertain that individuals of less than half the adult size may cast their cuticle twice within five months. This observation was made on eleven individuals which performed an ecdysis in April or May. These were isolated from the others and were observed to again shed the cuticle on different dates between June 7th and August 21st. The body-length of these individuals ranged from 6 to 95 cm. after the second ecdysis noted (the mean length of an adult being 2.1 cm.). The shortest time between two ecdyses observed was forty-three days (April 25th to June 7th). More than two ecdyses were not observed in a single individual, even among the several which were kept in captivity for about two years. It is, however, quite possible that an ecdysis occurred occasionally without being detected, as during certain weeks the animals were looked at only every other day. So far as these observations go, it will be seen that they confirm the statement of Cornelius that the post-embryonic development of this species is relatively long.

The Act of Ecdysis.-The mode of shedding the cuticle is described and illustrated by Miall and Denny', to whose account it may be added that the whole process usually lasts for two or three hours, though sometimes the crumpled mass of cast Op. cit. p. 32.

1

cuticle remains adhering to the hinder end of the abdomen for more than a day after. Coloration of the new cuticle commences as blotches on the dorsal surface and extends gradually to the outlying parts, the extremities of the limbs becoming completely coloured by the end of the third day after ecdysis. The animal remains unusually still during this period and eats very little, apparently not at all during the first day.

Loss of Appendages during Ecdysis.-Five cases were observed in which normal, and apparently uninjured, tarsi were broken off during ecdysis. This may result from a struggle to free the leg during its sliding away from inside the old cuticle. As already pointed out, the region where a break in the leg occurs most easily is the suture between the femur and trochanter, so that these occurrences of a break at a normally stronger point may perhaps be accepted as an indirect argument that there is a certain degree of autotomy in the usual rupture between femur and trochanter.

Note on the Numerical Proportion of the Sexes.

Among the whole number of adult specimens collected for the purposes of the present enquiry, the actual and percentage distribution of the sexes were:

As the habits of the two sexes are similar, the above is probably a fairly accurate indication of their usual numerical proportion, except possibly in the case of Phyllodromia germanica.

In a future communication I hope to give an account of the structure of the reproduced leg with special reference to the fourjointed tarsus, and also of certain cases in which apparently reproduced tarsi were found to be in a three-jointed condition.