8 THE FAUNA OF SCOTLAND. 



Blennocampa are not spiny. Those of EmjyJujtus, and other genera 

 of the group, are very similarly formed, and all bore into stems 

 to pass the pupa state. The larvae of Do! ems, Allantus, 

 Macropliya, and Tenthredo cannot be defined from eacli other. It 

 is thus perfectly clear that the forms of the larvae are far too 

 much of an adaptive nature to play an important role in the 

 definition of genera. 



We know too little of the embryology of the family to say 

 whether it is of any use; but judging from what is known about 

 the embryology of insects generally, I should say it will be of 

 little value, and indeed it appears to be of little importance in 

 defining and classifying the orders of insects. 



There is another character which has been employed in the 

 making of genera that requires to be considered. I allude to 

 secondary sexual characters. For example, in Cladius we have 

 three groups which are distinguished by the males having 

 differently formed antennae, the difference in the case of Cladius 

 difformis and C. padi being considerable, while the females of the 

 three divisions cannot be defined. In a case like this, I should 

 certainly say that genera ought not to be founded on structures 

 that are confined solely to one sex, no matter how conspicuous 

 such structures may be. 



There are other points, not perhaps of generic value, which are 

 nevertheless of use in indicating the affinities of particular species 

 or genera, — such as the manner of oviposition, of pupation, the 

 general habits of the larvae and imago, and their geographical 

 distribution. 



In connection with the classification it may be of interest to 

 consider for a moment the geographical distribution of the sub- 

 tribes. This will be best shown by the Table, in which the regions 

 and sub-regions are given after the masterly work of A. E. Wallace 

 on the geographical distribution of animals (see Table, page 5). 



From this table it will be seen that the Tenthredlna are nearly 

 confined to the Palaearctic and Nearctic regions, and when they 

 extend beyond them, e.g., into the Oriental, they are either 

 represented by Palaearctic genera, or by genera differing but 

 little from them, and which fall naturally into Palaearctic sub- 

 tribes. The true Cinibic'ides are also Palaearctic and Nearctic, 

 penetrating no further south than the Alleghanian sub-region of 

 the latter. In the Neotropical region there are three peculiar 



