35 



the larger upper facets giving a wider range and the lower ones a 

 more distinct view of close objects. A fact to be noted in connec- 

 tion with the prominence of the eyes is the extreme mobility of the 

 whole head, which is only attached to the thorax by a very slender 

 neck, thus giving scope for wide range of movement. There is an 

 extraordinary tropical family {Diopsidce) which has the side of the 

 head produced into two horns, each bearing an eye and an antenna at 

 the extremity. I recollect that when some specimens were being 

 exhibited at a meeting of the Entomological Society the question of 

 the use of this arrangement cropped up, but beyond the suggestion 

 that the fly was able to see on both sides of a leaf at once no reason 

 was brought forward for this strange modification. The antennae are 

 very variable within certain broad limits. In the Nemocera (daddies, 

 gnats, etc.) the antennae consist, as the name implies, of a varying 

 number of joints arranged in a thread-like form, the usual insect type ; 

 but in the other two great divisions of the order, the Brachycera and 

 the Atherkera, they are formed quite differently, there being, as a 

 rule, two short basal segments and a larger third one, the flagellum, 

 which frequently has a slender bristle attached, called the arista. 

 The arista may be plumose or bare, terminal, dorsal, or altogether 

 wanting, and the proportionate size of the three joints may also vary 

 considerably, with the result that much use is made of the antennae 

 in classification. The mouth-parts are variously adapted for the 

 purposes of suction. To take two types, I suppose all are familiar 

 with the enlargements of the mouth-parts of the house-fly and the 

 piercing apparatus of the mosquitoes. In the CEstridcB or bot-flies 

 the mouth-parts are atrophied in the imago. The palpi are of impor- 

 tance from the point of view of classification, and the proboscis, 

 though sometimes jointed and doubled back and sometimes perma- 

 nently extended, is never coiled as in the Lepidoptera. 



The legs are quite as variable as the antennae. They are found 

 modified for purposes of sexual ornamentation, especially in the 

 DolichopodidiB, furnished with strong bristles and spines in some 

 predacious Diptera, and with extremely delicate " touch hairs " in 

 other species. It is often possible to name the family of a specimen 

 from the legs alone without other aid. 



The body varies from the typical house-fly form to the globular 

 Cyriidce, the wasp-like Conopidce, and the flattened Picpipara ; it 

 may be bare and polished, covered with dense pubescence or with 

 coarse bristles, and numerous gradations exist between the different 

 states. The bristles and hairs of Diptera have already given rise to 

 a distinct form of classification known as "chfetotoxy" and are well 

 worth study, serving as they do for purposes of offence in the 

 predacious AsilidcB ; defence, possibly, in the dense, hairy covering 

 of some species ; cleanliness, as can be seen by anyone who watches 

 a house-fly cleaning itself; and ornament, as in some of the 

 DolichopodidcB. The wings of Diptera differ less, perhaps, than 

 any other portion of the external anatomy (I am not speaking now of 



^ 



