22O WHITMAN. 



kese, taking advantage of the headquarters selected by the birds. 

 If we were to take up some groups of migratory birds, we might 

 find it desirable to migrate with them, changing our location to 

 suit their summer and winter predilections. For forms settled 

 in tropical regions, longer excursions might be necessary, and 

 this might lead to the development of a new farm. These possi- 

 bilities do not in the least conflict with the plan proposed for 

 Wood's Roll. No matter where a farm be located, it must have 

 headquarters, and occasionally extend its field of work to more 

 or less distant points of interest. 



The headquarters must be in close touch with a laboratory, 

 and both should be in a place where the natural advantages and 

 the organization are such as to draw a large number of investi- 

 gators most emphatically not in a place that invites a large 

 number of spectators, as in the public parks of cities. In this 

 latter respect Wood's Holl is most perfectly adapted to the pur- 

 pose, and the prospect is good that we shall never be troubled 

 with throngs of summer visitors. This is an ideal feature in the 

 situation that it would be hard to duplicate. 



The practical question arises as to how to proceed with the 

 development of a farm. Our limited experience strongly con- 

 firms the opinion with which we set out, namely, that the best 

 method is to develop the farm slowly, section by section. Each 

 section should be a group of related species, selected with a view 

 to combining a wide range of problems. It should be developed 

 and directed by an investigator prepared to make it his life-work. 

 This investigator, or director, should have the support of a num- 

 ber of assistants competent to deal with special problems, one or 

 two artists, a photographer, a stenographer, a keeper and a busi- 

 ness superintendent. 



Developed in this way, the cost of maintenance would not be 

 heavy at first. Ten thousand dollars a year would support a 

 large and thriving section. The multiplication of sections and 

 the gradual growth of the work would call for a larger income. 

 A farm of ten large sections would require an endowment of a 

 million, and it is easy to see room for many millions. 



If the scheme here outlined approaches the ideal which science 

 is waiting to see realized, it will be seen that the farm does not 



