STATE HORTICULTURAL SOCIETY. 39 



less familiar kinds. It is not at all unlikely that among the little-known 

 denizens of the forest and field an actual knowledge would show us many 

 that we should do well to encourage by special protection, perhaps even 

 by artificial breeding. At any rate, every ornithologist knows that some 

 of our most useful kinds seek the protection and abundance of man's 

 neighborhood ; and, as these were, of course, originally wild and self- 

 dependent, it is not unreasonable to suppose that others may be brought 

 to this state of semi-domestication, if this is proved desirable. Besides, 

 until we know the manner and extent to which the different species com- 

 pete with each other for food at scant seasons, we cannot tell how best to 

 favor the increase of those which we find beneficial. 



We must also take account of the fruits and grains in bearing at each 

 season of the given year, in order that we may arrive at a knowledge of 

 \\\t preferences of birds. It is possible that a bird may feed largely upon 

 insects in the early spring, and yet eat only fruits when these are ripe. 

 We must know the kinds of food in considerable detail. If this is 

 vegetable, we must discriminate the kind of seed or other vegetable 

 matter, at least so far as to tell whether the bird is directly injurious or 

 not ; and if it is, we must know the kind of fruit or grain upon which 

 it feeds. 



If the bird is insectiverous, we must know at least the order of the 

 insect, apd must be, in most cases, able to say further that the insects 

 represented are or are not of certain beneficial families or genera. If pos- 

 sible to make out the exact species, this adds certainty and completeness 

 to the information, and makes it of special scientific value. If the bird, 

 like a hawk or an owl, feeds upon other birds or upon mammals, we must 

 know the kinds of these, in order to tell whether they are destructive or 

 not ; and, in general, the more accurate and detailed is our knowledge, 

 the more reliable will be our results. 



But all this, it may be said, involves a long and tedious research, 

 which it will take time and money and talent and learning to accomplish; 

 to which I reply, that these are just what valuable knowledge generally 

 cost. Whether in this case the probable results are worth the effort, you 

 are best qualified to judge. 



Of course, so exhaustive an investigation as this is not necessary to 

 the beginning oi accurate and valuable knowledge. We may begin to 

 gather the fruits of our labors from the first by a nearer approach to the 

 truth ; but nothing less than this will clear the whole field of doubt and 

 error. 



Three methods of observation have been employed in this matter. 

 The birds have been watched while taking their food ; they have been 

 fed in confinement, and the contents of their stomachs have been exam- 

 ined after death. The second method evidently shows only what the 

 bird will tdX in confinement of such food as is placed before him, and 

 furnishes no data which we can use with any security in making up an 

 account of his food in freedom, when he has to search for himself. Tiie 

 first shows us what the bird probably or possibly eats at the particular 

 time, and under the particular circumstances when noted. It lacks cer- 



