230 General Notes. \_x%, 



and voted for it then, and heartily applaud it now ; for, if not the most 

 convenient, it is the most logical and biological procedure to pass from 

 the 'lowest,' i.e., the most generalized forms to those which are the 

 most specialized, or 'highest'; such being apparently the 'natural' 

 course of evolutionary processes. I also think we did the business well, 

 on the whole; nobody doubts that our List passes from bottom to top of 

 the avian series, about as smoothly as the families could be arranged in 

 any single linear sequence — understanding, of course, that «o one linear 

 arrangement can possibly be natural, yet that some one such is a 

 mechanical necessity of book-making. 



Granting then, that we turned the series of orders and families hind 

 part before in the best possible manner, or at least in a manner free from 

 obvious objection, a very queer inconsistency crops up in our treatment of 

 the contetits of the numerous families. The same rule of reversal should 

 of course have been applied to the genera and species of each famih'. 

 But in point of fact such rule was not applied, in all instances at anv rate. 

 To put the case in a nutshell, we turned the list of families hind part 

 before, but generally left the sequence of genera in each family as they 

 had been in the previous lists I have named, which were modeled on the 

 high to low principle. That this is a fact, anyone can satisfy himself by 

 inspection of our Check-List ; but the agitated searcher for light on this 

 point may have to go througli the whole of our work, before the full 

 magnitude of our offence dawns upon him. I will put him on the track 

 by citing a single case of what I mean, and he may follow up the investi- 

 gation to any extent he pleases. 



In the family Anatidae, the general treatment of the subfamilies and 

 genera had been, in those lists which went on the high to low order, 

 to begin with the Swans or Geese, Anserinse or Cygninae ; go on to the 

 Anatinse, which in fact inosculate with Anserinae through the Shell- 

 drake group, etc.; pass thence to the Fuligulinae ; and finish with the 

 Merginje. We use these identical subfamilies, and, as I think, advisedlv ; 

 we also have after a fashion reversed their sequence, so that my criticism 

 is to some extent weakened in this verj' case. But no one doubts the 

 specially close connection of Merginte with Fuligulime ; assuredly these 

 two subfamilies should come together. Instead of that we begin, cor- 

 rectly, with the Merginie, as the ' lowest ' members of the family ; then 

 jump directly to the Anatinoe ; begin the Anatinse with Anas, at the top 

 of the list, and run the gamut of its genera from ' high to low,' in the 

 good old-fashioned way; put Aix far from its obvious and undisputed 

 position; pass on to Fuligulin;e and run do-vn that list of genera to 

 the Scoters, Eiders, and Erismaturine genera; whence we jump again with 

 admirable agility but questionable propriety to the Geese proper, Anser- 

 inse, and so on to the Cygninae. 



I am not here raising any real taxonomic question. I assume that we 

 are in substantial agreement of opinion as to the natural relationships of 

 the subfamilies of Anatid£e, but contend that their sequence in our List 



