CABINET FOR SKINS. 109 



the smallest birds, a multiplicity of shallow drawers, accom- 

 modating each one layer of specimens ; it is better to have 

 fewer deeper drawers, into which light shallow movable trays 

 are fitted. These trays never need be of stuff over one-eighth 

 or one-fourth thick, and may have stiff pasteboard bottoms 

 glued or tacked on. They may vary from one-half inch to two 

 inches in depth, but this dimension should always be some 

 factor of the depth of the drawer, so that a certain number of 

 trays may exactly fill it. They should be just as long as one 

 transverse dimension of the drawer, and rather narrow, so that 

 two or more are set side by side. Finally, though the}' may 

 be of different depths, they should be of the same length and 

 breadth, so as to be interchangeable. They may simply rest 

 on top of each other, or slide on separate projections inside the 

 drawer. Such trays are extremely handy for holding partic- 

 ular sets of species, to be carried to the study table without 

 disturbing the rest of the collection. 



If a collection be so extensive that any particular specimen 

 may not be readily hunted up it will be found convenient to 

 have the drawers themselves labelled with the name of the 

 group within. A collection should always be methodically 

 arranged preferably according to some approved or sup- 

 posed natural classification of birds ; this is also the readiest 

 mode, since, with some conspicuous exceptions birds of the 

 same natural group are approximately of the same size. If I 

 were desired to suggest proportions for a private cabinet of 

 most general eligibility, I should say four feet high, by three 

 feet wide, by two feet deep, in the clear ; this makes a portly 

 yet not unwieldy looking object. It is wide enough for fold- 

 ing doors, to be secured by bolts at top and bottom, and lock ; 

 not so high that the top drawer is not readily inspected ; and 

 of proportionate depth. Such a case will take seven drawers 

 six inches deep either of the full width, or in two series with a 

 median partition ; these drawers will hold anything up to an 

 eagle or crane. A part of them at least should have a full 

 complement of such trays as I have described say three or 

 four tiers of the shallower trays, three trays to a tier, each 



