o(5 INSECTA MADERENSIA. 



thorax has altered considerably in shape, being much narro-\ved behind and punc- 

 tured, and the eMra are nearly always free. In this state it continues for about 

 1500 feet, when again emerging into the broad daylight of the open hills, it 

 recommences to mould itself as it did below ; until, having reached the summits 

 of the loftiest peaks, more than COOO feet above the sea, it has almost (though not 

 entirely) assumed the features which characterized it on the shores beneath. Tliis 

 is of coiu'se only a general account of the changes which take place during its 

 upward progress ; yet, although exceptional cases, as to every other rule, "nill now 

 and then Ijc met with, I beHeve it to be perfectly true on a large scale. That 

 modifications of a slightly different kind occm- at parallel altitudes in the other 

 islands of the group is also certain ; but we must not forget that the disappearance 

 of the dense forest everywhere except in Madeu-a proper may have re-adjusted for 

 those particular spots the law which there also in all probability once ol:)tained. 

 ^Moreover, whatever the caiises may be which operate in these remote spheres to 

 affect the insect life wliich has 1)ecome isolated upon them, it is certain, from 

 observation, that theu* working is not accidental, but depends on the peculiar 

 circumstances of the respective localities, since species of even opposite natiu'es 

 are affected in a smiilar manner on the same rocks. Thus we accordingly find, in 

 tlie present case (as in others which I have ah'eady liad occasion to comment U2)on 

 under the genera Scarites, Uiiri/r/inft/uis, Calathus and Olisthopus), that the speci- 

 mens wliich have been detached on the tAvo northern Dezertas have attained a 

 larger size than those on any other island, that the Madeiran ones are the smallest, 

 whilst those in Porto Santo are unusually depressed. Although confined to the 

 sylvan districts and less abundant than any of the remainder, I have nevertheless 

 assiuned var. a. to be the normal state ; fu"st, because, judging from the description, 

 I believe it to be the particular form descri])ed by Dejean in 1829, and secondly, 

 because the ]\Iadeira Islands, as theu* name implies, being by natm-e islands of 

 wood, it is the variety in all probability which would have formerly predominated 

 throughout the group, — the other modifications being the result in some measm'e 

 of the destruction of the timber, and partially therefore, though indii'cetly, refer- 

 able to the agency of man. I have contented myself, in the above diagnosis, by 

 indicating three varieties only, since it is impossil)le to define the limits and ranges 

 of subsidiary modifications, my sole object having been to give a generally correct 

 idea on a broad scale. 



This species, which, in all its phases, is well distinguished from the rest of the 

 genus inhabiting these islands, presents a type of structure peculiarly its own. 

 Thus, for instance, it is universally apterous (whereas the others are winged), the 

 nienlum lias no tooth in the centre of its einargination (a structure of very rare 

 occm-rence in the trvie HarpaU), and the extreme apices of each of its elytra are 

 always truncated ol)liquely, causing them to diverge minutely, or divaricate, at 

 the termination of the suture. But perhaps its most singular character, and in 

 which it (lilTcrs from every other Ilarpalus with which I am acquainted, consists 



