BY K. J. TILLYARD. 95 



underside of each of the inferior lateral ones, are fringed 

 with stiff hairs. When the three plates are in sitti, I suppose 

 that these hairs must, in some manner, interlace to form the 

 necessary filter, keeping the rectum itself clear of all impuri- 

 ties in the water. When watching the larvae breathing in 

 muddy water, I noticed particularly that these plates are 

 drawn so closely together as to make the anal end appear 

 almost closed during inspirdfion , yet, during exjnration, the 

 rectum is wide o-pen, just as wc find it in the exuviae. This 

 seems to suggest that the full-fed larva, when about to 

 emerge, and as soon as it has climbed above the water-level, 

 keeps its rectum open for the purpose of air-hreathinri . If 

 this be the case. Dr. Ris' conjecture that the larva is an air- 

 breather, is seen to have some foundation in fact, but the 

 statement needs to be modified, viz., the larvae are only air- 

 hreathers when above water-level. I must confess that I fully 

 expected the larvae, when placed in muddy v/ater, to push 

 their anal ends up into the air to breathe, and I was rather 

 sm-prised to find that they preferred the muddy water to 

 anything else. Still, a little reflection will show that, even 

 if they keej) their channels of communication fairly open and 

 clean, yet these flimsy strictures must often collapse, and 

 during periods of heavy rain it m.ust be quite impossible for 

 the soft, horizontal portions to hold together. I conclude, 

 therefore, that the larva do iisiially breathe muddy water, 

 which is filtered as described above, and that only on special 

 occasions, namely, when foraging above water-level, or 

 during the final ecdysis, do they breathe air. 



There is one more point of interest to determine. Do the 

 larvae take more than one year to reach maturity ? As far as 

 I know, practically all the imagines emerge in a period cover- 

 ins about the latter fortnig^ht of November. Now, of the 

 seven larvae taken at Medlow, three were practically full-fed, 

 with wing-cases of greater or less size ; but the other four were 

 considerably smaller, only, one might say, about half-grown, 

 and shewing as yet no trace of wing-cases. It seems, there- 



