1919.1 85 



It was j)(is.sil)le to olisel'w in some detail, iuuIlt ;i strong lens, tlie 

 inovenieiits by wiiieh the larva effeeteil its attaelnneiit to the grass-i-oots : 

 these are jjerhaps worth deseriliing, but in order to tlo so it will be iieees- 

 sary to give soiuu aecount of the strueture of the breathing-tube. '^I'his 

 organ is shorter than in must other eulieine mosquitoes, and far moi-e 

 eomplex. [t consists of broader basal portion (corresponding to the air- 

 tube of Culex) and a narrower, black apical portion. This latter is quite 

 olniously homologous with the valvular closing apparatus of the ordinary 

 culiclne larva, though this fact does not appear to have been recognized 

 by previo\is writers on this group of mosquitoes. In TaciiiurliiincIiKs 

 the dorsal jjairof vanes (always smaller than the ventral) is considerably 

 reduced in size, but the stout cur\ed In-istle which they carry is, on the 

 other hand, ver}' strongly developed. The ventral pair of valves are large 

 and elongate, and form the black apical ijortion of the air-tube ; they are 

 joined together by membrane, and in this way are modified to form an 

 immovable sheath for the protection and support of the complex internal 

 organs. These last comprise a flat piece with a saw-like dorsal keel, 

 and a pair of sharply pointed pieces on either side of it, attached to the 

 end of the common hollow space formed by the conjoined tracheal tubes ; 

 also a ])air of more vcntrally phiced pieces, each with two very strongly 

 chitinized hooks at the tip, hinged on to a stout axial rod,. which again 

 is fixed to the junction of the tracheal tubes. Finally, there arc three 

 pairs of sharp hooks in the membrane at the apex of the ventral valves. 



The breathing-tube of Majisunioidcs aj'ricaiius, well described and 

 ligured by Ingram and Scott Maclie (Bull. Eiitora. liescarch, viii, 1*J17, 

 pp. 137-140 j, has a closely similar strueture, only differing in certain 

 minor details, such as the general shape of the air-tube and of the 

 ventral valves, and the number of teeth in the saw (about 1-1 instead 

 of about 7). 



When the larva approaches a rootlet with the object of affixing 

 itself, which it does tail-foremost, the first movement which can be 

 observed is that the strong biistles on the ventral valves are waved 

 about, and their sharp tips inserted into the spot chosen. Doubtless, as 

 suggested by Ingram and Maetie, these bristles are sensory, and may 

 " transmit a stimulus enjoining the larvae to attach themselves.'" The 

 tips of the ventral valves are then bi-ought into contact with the root, 

 and the three pairs of hooks, which are normally inside the tips of the 

 valves, ai'e everted. These hooks are directed posteriorly, the points of 

 the stout bristles anteriorly, so that between them they can obviously 

 afford the larva a tolerably firm preliminary " foothold." The hooks, it 

 should be said, act quite independently of the other internal structures. 



