172 CRUSTACEANS. 



evidence, that the shell of the Cancer pagiirus may be repeatedly gene- 

 rated with scarcely any deviation from its original quality. 



The first drawing, it wiU be observed, had been executed on the 

 3d of November, fig. 1 ; and the last on November 10, or with the lapse 

 of above a year. — Fig. 4. 



But the animal continued stiU growing, and it cast its shell for the 

 fourth time on the 29 th of March following, in the second year of its 

 captivity, or above 500 days after the period of the first delineation. Yet 

 no colouring matter was visibly incorporated with it. 



The under surface of the empty shell, cast on March 29, was ad- 

 mirably and faithfully delineated after the new animal had left it. — 

 Fig. 5. 



But a singular fact was betrayed on this last exuviation. The new 

 specimen, as I must still call it, now produced from figs. 4, 5, was de- 

 fective of the whole eight limbs. It had a strange appearance without 

 them. Only the two claws, both tipped with black were present, the 

 other eight were totally deficient. The rest of the other parts, all ex- 

 cept the tips, were beautifully white. 



I did not at first observe this extraordinary deficiency, the creature 

 was so very tame and tranquil ; it rested on the breast, but being acci- 

 dentally reversed it exposed the truth. The drawings now made, fig. 6, 

 back, fig. 7, breast, may be compared with those preceding. 



In a few days the mutilated animal fed copiously, and continued to 

 do so. It seemed quite well, and remained very tame. Its only pecu- 

 liarity was falling frequently on its back, from inability to balance itself, 

 for want of so many limbs. 



I preserved this creature with uncommon care, partly from indulging 

 anticipations of futurity, partly from curiosity, to determine whether my 

 various conjectures above alluded to were sound or erroneous. 



Animal life is so fleeting, external form so easily impau-ed, I was 

 continually on the watch. 



Yet these animals are well adapted for experiment, being fed and 

 preserved with such facility. 



After the defective shell investing the living creature, figs. 6, 7, had 



