HOUSE FLiES AND BLUEBOTTLES 185 



surface as is required by the hypothesis. The mechanism 

 which brings about these curious results must rather be 

 sought for in the hairs themselves. The pads are, in 

 fact, hollow, and contain, protruding into their cavity, 

 the nipple-shaped ends of a sac which occupies more or 

 less of the interior of the last four tarsal joints. This 

 sac secretes a perfectly clear viscid liquid, which exudes 

 into the pad, and from that into the hairs which project 

 from it. These hairs, which are said to number about 

 1200 on each pad, are hollow, terminate in tubular 

 orifices, and are kept full of the secretion. Hence the 

 entire surface of each pad is crowded with a number 

 of viscid points; and as there are in all twelve pads, 

 two to each foot, these, when applied to the surface 

 over which the fly is walking, produce an adhesion 

 sufficiently strong to support the slight weight of the 

 insect. The viscid liquid soon hardens on exposure to 

 the air, but no doubt remains liquid while covered by 

 the pad. Thus the insect is, as it were, at every step, 

 temporarily glued to the surface over which it is travel- 

 ling, and leaves on a clean surface little rows of dots 

 as its footprints. This does not necessarily involve any 

 violent strain in wrenching the foot off again, since the 

 tarsus is raised obliquely, and each row of hairs is 

 therefore successively detached, somewhat in the same 

 way as a piece of court plaster may be easily removed 

 from the hand by taking it up at one end and raising it 

 obliquely, though it might resist a considerably greater 

 strain if merely pulled at one end without raising, all 

 the points of contact then combining to resist the 

 strain. 



There yet remains the head of our bluebottle to be 

 examined. It is a broad almost triangular plate, united 

 to the thorax by a junction so slender and delicate that 



