TRANSFORIMATIONS OF THE GNAT. 605 



time, and any number can be taken for the purpose of experi- 

 menting. Their shape very much resembles that of the life- 

 boat now in use, and, like the life-boat, the egg-boat cannot be 

 sunk, and if capsized rights itself again immediately. Even if 

 some of these boats be placed in a vessel of water, and the^ 

 contents of the vessel be poured from a height into the pond, / 

 the little boats float at once to the surface like so many corks, 

 and each, as it rises, assumes its proper position. 



In due course of time the larva is hatched, pushes off the 

 lower end of the egg, which opens like a little circular trap- 

 door, and allows itself to float off" into the water. The larva is 

 a quaint-looking little being, with a long body terminated at 

 one end with a large round head, and at the other with a 

 forked tail. When examined through the microscoi^e the larva 

 is a most curious creature, the semi-transparency of the body 

 rendering the internal organs almost as plainly visible as if 

 there were no skin at all. The young and small larva}, which 

 have just shed one of their successive skins, are better for 

 microscopical examination than those of a larger size, because 

 their integuments . become more opaque with age. Through 

 the centre of the body the digestive organs are marked by their 

 darker hue, and just above them pulsates the 'dorsal vessel' 

 which stands insects in the stead of a heart. 



On either side of the body runs a rather dark tube, and the 

 two, joining each other at the angle of the fork of the tail, 

 turn off to one of the points of the fork, and run side by side . 

 along it. These tubes are the two principal canals of the 

 respiratory apparatus, as is easily seen by putting a high power 

 to the microscope — say, an object-glass magnifying some two 

 hundred diameters. When this is done, the spiral thread which 

 is coiled round the breathing tube of insects becomes plainly 

 visible, and at once declares the character of these dark vessels. 

 At reoular intervals smaller vessels branch off to the other 

 parts of the body, and the tubes are finally lost in the com- 

 paratively opaque head. 



After shedding the larval skin several times, the pupal state 

 is assumed. In this condition the future Gnat can move about 

 with some activity, but it cannot take nourishment, all tlie 

 apparatus of its mouth being enveloped in the pupal skin. 

 After passing a short time in this state, the pupa cracks along 



