23 



geschwindigkeit infolge Gasdiffusion durch die Tracheenwande wie bei Corethra 

 fmdet bei Cu/e;r-Larven nicht statt". 



Generally these statements are most probably quite correct ; many of the experi- 

 ments which KOCH has carried on, I have occasionally had an opportunity to ob- 

 serve. It is quite right that the larvae of many mosquitoes, when they leave the 

 bottom for the surface, rise slowly in a vertical position without any motion; 

 the head is always downwards; that the sipho plays the role of "Schwimglocke" 

 is most probably quite correct; many species, f. i. C. morsitans, can often be seen 

 in quite the same manner sinking from the surface down to the bottom; but here 

 the head is also directed downwards and not upwards as KOCH supposes; in this 

 case the sipho has no significance as "Schwimglocke". With regard to 0. rusticus 

 I have very often seen the larva with an air-bubble at the top of the sipho rise 

 vertically from the bottom and without making any actual movement reach the surface. 

 But of this species I have also seen larvae, which first passively sank downwards and 

 then, before they reached the bottom, stood still for a moment in the water layers and 

 then again passively rose to the surface. I have only observed this in this same larva; 

 that the tracheal system in this case played a hydrostatical role is in my opinion un- 

 questionable. I am inclined to suppose that the mosquito larvae, with regard to the use 

 of their tracheal system as a hydrostatical apparatus, differ from each other. A more thor- 

 ough exploration of the tracheal system of the different mosquito larvae and the greater 

 or less use of it as a hydrostatic apparatus would probably give good results. It may 

 further be pointed out that the most elaborate tracheal system is found in those larvae 

 which live under bad respiratory conditions (tree-holes, polluted water, on the bottom 

 with the sipho pierced into the tissues of water plants, Tceniorhynchus). In these 

 larvae the tracheae are more band-like, especially in the thorax, the meaning of which 

 may probably be, that tracheae of this structure are able to store up larger amounts 

 of air than those whose transversal cuts are circular. We shall here only deal with 

 the highly remarkable tracheal system of the Tceniorhynchus larva. 



Between the thorax and the first abdominal segment are situated two large 

 bladders. At the first glance these bladders resemble those of the Corethra larva, but 

 differ from these in being connected with the two main tracheae by two smaller, 

 rectangularly curved tracheae. The point of insertion is situated at the line between 

 the thorax and the abdomen. Fig. 12 and 13 Tab. XV illustrates the description. It 

 will be seen that the front parts of the bladders are furnished with some few, short, 

 slender tracheae, and that the two large longitudinal stems of the body are broad, 

 band-like in the thorax and the first part of the abdomen. In fact these main 

 tracheae are double, their double derivation being still indicated by a longitudinal 

 fold; in the thorax each of the two trunks is further divided by another fold. The 

 result is two broad, flat, band-like bodies, terminating in front in three tracheae, the 

 first of which has a lateral course, whilst the second is directed forward and sends 

 branches to the head; the third, inner, branch is smaller and communicates with 

 that from the other side. In the abdomen each of the main tracheae sends a smal- 



