BY R. J. TILLYARD. 281 



present in the ancestral form. In such cases, a rigid application 

 of the law would result in a complete inversion of the phylo- 

 genetic conclusion. 



Meyrick liimself applied his law chiefly to the problem of 

 wing-venation in the Lepidoptera. In a case like this, where 

 almost the whole course of evolution is towards reduction and 

 simplification, tliere is not much danger in its use. Yet even 

 here caution is necessary. It is not true, for instance, that 

 veins which have once been lost cannot be re-developed; nor is 

 it true that new cross-veins cannot be formed in a wing-area in 

 which no such structures existed in the ancestral form. One has 

 only to examine the tracheation of the pupal wing in the Lepi- 

 doptera to see that there is an immense field of possibilities in 

 both these directions, owing to the persistence of the finer 

 tracheae, ramifying in all directions beneath the wing-membrane. 

 As long as these tracheae persist, so long must the possibility of 

 an imaginal venational meshwork reappearing be held to exist. 

 Moreover, in a case in which the wing is changing its shape, so 

 that a certain part of it may become broadened, it is exceedingly 

 likely that one or more of the tracheae underlying the veinlets 

 of this area will become lengthened and strengthened, so that it 

 may eventually lead to the replacement of an original short vein- 

 let by a longer oblique branch, which, on Meyrick's interpreta- 

 tion, would have to be regarded as an original archaic branch of 

 the main vein from which it springs One of the best examples 

 of this is the effect of the widening of the costal area in the 

 Psychopsidce, in which an originally fairly simple series of 

 veinlets has become greatly lengthened and enlarged, most of its 

 units branching many times; and all of them connected together 

 by newly developed series of cross veins, which were certainlv 

 not present in the ancestral form. 



The above example shows us that, even in so restricted a study 

 as that of Wing- Venation, Meyrick's Law can only be used with 

 great caution. It must be restricted to areas of the ivirtg that 

 are undergoing reduction. In the present state of evolution of 

 the Insect-Wing, it is certainly true that the tendency is towards 

 reduction in the great majority of cases. But this cannot always 



