944 MR. H. H. BRINDLEY ON REPRODUCED [DeC. 13, 



Heineken (46) has a single case o£ the antenna of Heduvius 

 being regenerated with 3 joints instead of the normal 4 after 

 mutilation in the " pupal " state. 



Lepicloptera . — Observations on the regeneration of appendages 

 in this order have yielded results so varied that it is not possible 

 to place them as a whole under either (a) or (b). It has already 

 been pointed out that much more experimental observation, and a 

 clearer understanding of the exact relations borne by the larval 

 and pupal appendages to those of the imago, are necessary before 

 the phenomena of reproduction in this order can be interpreted 

 satisfactorily. The effects of mutilation of the legs of larvse on those 

 of the imago were first investigated by Reaumur (71). and since 

 his time the subject has attracted only occasional notice. Prom the 

 small amount of recorded work it is not possible to gather how far 

 the results of a particular kind of injury at a particular stage in 

 the life-history are uniform. The largest number of experiments 

 on a single genus appear to be those of Newport (61) on Vanessa 

 larvae. In the imagos there was much variation in the condition of 

 the injured limbs. In all cases femvu', tibia, and tarsus could be 

 distinguished, but the number of tarsal joints varied considerably. 

 In all, however, the terminal claw of the tarsus was present. 

 This fact and the drawings which illustrate his paper suggest that 

 the reproduced tarsus in all these cases should be regarded as 

 representing the whole of the normal tarsus, rather than for 

 instance that a 3-jointed tarsus should be considered as equivalent 

 to three particular joints of the normal tarsus. However this 

 matter be looked upon, it remains that the tarsus is sufficiently 

 represented to bear the normal termination, the claw : so that these 

 observations on Vanessa are of particular interest as evidence of 

 the tendency, so characteristic of the instances quoted under (b) 

 above, towards the production of the terminal structure normal 

 to the limb, so that though normality in the number of joints may 

 be wanting, its actual termination is of normal structure. But 

 this tendency is not displayed in Lepidoptera with the constancy 

 it possesses in other orders, for the experiments of Melise (58) on 

 Sericaria and of Watson (77) on Dicranura gave results contrary 

 to Newport's as regards the tarsal claws. The total number of 

 observations by these two authors was, however, too small for a 

 fair comparison with those of Newport. 



But setting aside the Lepidopteia, it seems that we have at 

 least some clear indication that in other groups of Arthropods in 

 which reproduction has been studied there are two well-defined 

 types of structure assumed by the growths which replace lost 

 appendages, and that in any given case these two kinds are not 

 interchangeable. The reproduced limb either resem.bles the normal 

 in the number and conformation of its joints, or else it does not 

 do so. Now the interest of the cases in which the reproduced 

 appendage differs from the normal is that the various examples 

 exhibit a considerable degree of uniformity in their abnormal 

 characters — a uniforn)ity which is sufficiently marked to enable us 



