17 
The method of its use is as follows: given a bird in the hand of un- 
known species to determine its name. It is first compared with the first 
heading in small capitals numbered in Roman numerals—1, FEET FULLY 
WEBBED. If this description does not fit the bird, the next Roman 
numeral heading is referred to—1I, FEET PARTLY WEBBED, Or III, FEHT 
WITHOUT PRONOUNCED WEB. Assuming that the latter correctly describes 
the bird under discussion, we refer to the headings of next lower rank, 
which are numbered alphabetically with capital letters, where we find 
the alternatives—F, Legs long, and G, Legs short. Sometimes it may be 
difficult to decide whether a leg should be regarded as long or short, and 
the various pictured details following may then assist determination. In 
this case the legs we decide are not remarkably long, no longer in comparison 
to size of the bird than are the legs of a chicken or sparrow; we, therefore, 
under G, refer to a number of subordinate alternatives, distinguished by 
small initial letters—k, feet chicken-like, strong and compact for scrateh- 
ing; 1, feet strongly clawed for holding prey; m, feet small and weak; 
n, feet small or medium-sized, solidly made and legs covered with horny 
seales or plates. Glances at various feet shown under each heading will 
assist in determination. Assuming a decision in favour of the last, we 
compare our specimen with the next alternatives, numbered with ordinary 
Arabic numerals—19, two toes in front; 20, three toes in front. There 
can be little confusion here and we assume that our specimen having three 
front toes is one of the great body of perching birds. We, therefore, 
compare it with the following line detail drawings to see with which it 
agrees most closely. The bill is not wide and flat; it is, therefore, not a 
flycatcher; there are no ear-tufts or long hind toe and the nostril is not 
covered with feather tufts, therefore it cannot be either a Horned Lark, 
a Crow, or a Jay. The next picture, the Bobolink’s bill, catches our eye 
and the sparrow bill in the next lot. A glance through the remainder 
shows that our bird must be a bobolink or one of the sparrows. The picture 
p. 247, and description of the former, is nothing like it; therefore, we turn 
to the sparrows, read the general sparrow description, and remarks on p. 161, 
and then work through the pictures. After looking at all the illustrations 
we find that our specimen agrees with that of the Song Sparrow, and on 
reading over the distinctive characters we have our opinion confirmed. 
It has the sharply striped breast aggregated in the centre, and is without 
either the yellow stripe over the eye of the Savannah Sparrow or the 
white outer feathers of the tail, as in the Vesper. We are, therefore, 
confident that, starting with no other ornithological knowledge than that 
the specimen was an Eastern Canadian bird, we have been able to refer it 
to its proper species. 
