28 



between life conditions and larvae structures on yet another point. In the bottom 

 dwellers we find a remarkable development of special hairs on the apical part of 

 the sipho which are either missing or only but slightly developed in the surface 

 dwellers. These hairs play a certain role in the life of the bottom dwellers, providing 

 them with points of support when they rest on the bottom with the dorsum down- 

 wards, or hang down from water plants. 



b. The pupa. 



It is a well-known fact that the pupa stage of the insects is mainly a resting- 

 stage. In total repose, without any supply of food and often in darkness, the great 

 alterations in external and internal anatomical structures are now to be accom- 

 plished. The power of locomotion in the pupa stage is therefore commonly very 

 greatly restricted. Many pupae are known to be quite unable to change their place. 

 Some of them, more especially those w r hich live in water, are able to make 

 respiratory movements, oscillations with the abdomen (pupae of Chironomidce a. o.). 

 The real power of changing locality is, if present, commonly restricted to the last 

 days of the pupa life. At this time these pupae push themselves out of their holes 

 and corridors to be as near as possible to the air and the sun when the last ec- 

 dysis takes place. 



In the Culicidce we apparently find the most movable pupae stages known in 

 the whole animal kingdom. Strong power of locomotion demands a supply of food ; 

 it might therefore be expected that the mosquito pupae would be able to take food; 

 this is however not the case. The appendages of the head and thorax of the future 

 fly is wholly enclosed in a common chitinous covering. This apparent physio- 

 logical contradiction is in accordance with two facts which probably have not 

 hitherto been sufficiently estimated. Firstly the pupa stage is extremely short; in the 

 tropics there are species e. g. Psorophora which, according to HOWARD, DYAR and 

 KNAB, pass the pupa stage in a shorter time than one day, and even in our lati- 

 tudes it does not regularly last more than a few days. In periods of very cold and 

 rainy weather the pupa stage can be prolonged beyond a week or more, but this 

 is against the rule. Only once have I in my laboratory seen the pupa stage pro- 

 longed beyond three weeks; these pupae belonged to the last broods of C. pipiens, 

 taken as larva? into the laboratory on October the fifth; pupae arriving on October 

 the tenth, lived even on the fifth of November; a few reached the imago stage, but 

 the others died. The temperature of the room was not more than 10 C. and for a 

 week always less than 5 C. 



Leaving a more thorough description of the mode of locomotion for a later 

 paragraph, we will here restrict ourselves to the following remarks. It may gener- 

 ally be maintained that we have highly overrated the power of locomotion. 



HOWARD, DYAR and KNAB (1912 p. 98) write as follows: "The pupa is active 

 and capable of moving rapidly through the water" further: "The pupa is very easily 

 alarmed and at the appearance of a shadow or at a slight disturbance of the water, 



