BY R. J. TILLY ARD. 675 



done before February, these larvae must at any rate have been 

 liatched by March. So that they are wow fifteen months old. 



As regards resistance to starvation, these larvae went without 

 food for about three months, and I have no doubt could endure 

 a longer fast if necessary. Often, in the winter, their mountain 

 haunts are frozen up, and they are compelled to go without food. 

 As regards 5. macrostignui, which inhabits the thick mud at the 

 edge of a mountain bog, it is probable that it is often completely 

 enclosed in the hard cake of frozen mud in which it lives, with- 

 out being any the worse for it. 



In conclusion, I find that, as far as my experiments go, most 

 dragonfly larvse can go without food for long periods. I suspect 

 that many others besides those which were made the subject of 

 these experiments are able to exist as larvae for two or even three 

 years. I do not, however, think that many can resist drought; 

 that is a power possessed only by a few Libellulidce, and possibly 

 also by Petalura and Argiolestes. 



Postscript (added 18th October, 1910). — Since thi.s paper was 

 read, much that is of interest has occurred with regard to the 

 unknown Libelhdid larva mentioned therein. By the middle of 

 September it appeared quite full-grown, and on several successive 

 days it moved up the twig and rested partly out of the water. 

 However, the weather not being very warm, it returned again to 

 the bottom of the jar, and became very listless and inert. On 

 October 7th, it emerged from the water, and climbed well up the 

 twig, resting some three inches above the surface of the water. 

 There it remained all day, unable to effect the final change. The 

 weather wa« rather cold and exceedingly windy, so that it was 

 unfortunately by no means a good day for a successful emergence. 

 On my return home I found that the larva had fallen off the 

 twig dead. Its thorax was much swollen, and the parts of the 

 head were dry and partly cracked; so that it was evident that 

 the final change had begun, but that the larva had lacked strength 

 to complete it. How far this failure was due to unnatural con- 

 ditions I cannot say, for the larva was quite strong and healthy 

 until a few days before it died. 



