INTERMEDIATE AND ABERRANT GROUPS. iat 
there are many difficulties in the way of their practical definition. This 
is partly because all birds are singularly inter-related, presenting few broad, 
unequivocal, unexceptional characters in the midst of numberless minor 
modifications, and partly because the higher groups, no less than species and 
varieties, shade into each other. In our illustration, for example, we find 
exactly intermediate aquatic machines; thus, it would be difficult for a 
landsman to say whether an hermaphrodite brig belonged to the ship family, 
or the schooner family ; he would have to decide according as he considered 
number of masts, or shape of sails, the more essential family character. 
But the tntermediate groups which remain to be examined are not of this 
ambiguous nature ; they are unequivocally referable to some particular group 
of the next higher grade, and, being subordinate divisions, they are distin- 
guished by the prefix sub, as sub-order, sub-family. Though somewhat 
difficult to define, they are, I think, susceptible of intelligible, if not always 
precise, definition. A sub-group of any grade is framed, without taking 
into consideration any new or additional characters, upon the varying prom- 
inence of one or more of the characters just used to form the group next 
above. In our formula above & (abc) for a certain family of the order a, 
suppose the family character a to be emphasized, as it were, and to pre- 
dominate over } and c, to the partial suppression of these last: then a sub- 
family of x (abc) might be expressed thus:—«# (Adc) ; and it is further 
evident, that there will be as many sub-families as there are groups of birds 
in the family representing varying emphasis of a, or 6, orc; asx (a Bc),% 
(Ab C), ete. While we take account of new characters of another grade, 
in forming our successive main groups, in our sub-groups, then, we recog- 
nize only more or less of the same characters. But the distinction is not 
always evident; nor is it observed so often as, perhaps, it should be. 
§ 24. Typican anp ApeRRANT Groups. Waiving what might be rea- 
sonably argued against considering any group specially “typical” of the 
next higher, we may define a convenient and frequent term: —The typical 
genus of a family, or family of an order, is that one which develops most 
strongly, or displays most clearly, the more essential characters of the next 
higher group, of which it is one member. And in proportion as it fails to 
express these in the most marked manner, either by bearing their stamp more 
lightly, or by having it obscured or defaced by admixture of the characters 
of a neighboring group, does it become less and less typical (“ subtypical ”) 
and finally aberrant. Suppose the ordinal symbol «, as before, to represent 
the sum of various ordinal characters, more or less essential to the integrity 
of the order; then obviously, the family characters abc, or def may be com- 
bined with a varying value of w; thus, a’ (abc) or « (def) and the formula 
of the typical family would be a" (a—f). Thus, it is characteristic of most 
thrushes (Turdid@) to have the tarsus booted, but all do not have it so; 
therefore, in subdividing the family, we properly make a division into 
thrushes with booted tarsi, and thrushes with scutellated tarsi; the former 
are typical of the family, the latter sub-typical or even aberrant. 
