Vol. XI. 

 191 1 



1 Correspondence. ^39 



a method is needed that w\\\ reveal it withoid unnecessary 

 mental exertion. A most important item is tiic systematic 

 descrijition and measurement ol sets of eggs, where!)}' an accurate 

 contigiuation of their various peculiarities is conveyed to students. 

 This attained, ornithologists will be furnished with material which 

 will help in the elucidation of the laws which govern the multi- 

 tudinous variations, which at present are very imperfectly under- 

 stood. The terminology needs to be more definite to meet the 

 requirements of expanding research. With the present system, 

 I venture the opinion that very few cast more than a casual glance 

 at measurements given. In ^Hsjilacing old methods of science, 

 the new must justify itself by obvious advantages. The method 

 I suggest is a division of the egg into definite areas, so that de- 

 scription and examination may be facilitated. It is analogous 

 to that employed by astronomers in dividing the surface of the 

 moon into definite areas, each of which may be surveyed without 

 reference to the contiguous ones. The system may also be 

 likened to the principle of geographers of animal life, who divide 

 regions into sub-regions to simplify their tabulation. It is there- 

 fore necessary to have some kind of table to produce statistical 

 evidence wherewith to arrive at the mean shape, size, and colour 

 of eggs of any given species. Every oologist has experienced 

 difticulty in identifying, or discriminating between, eggs of allied 

 species, and any attempt to formulate a rule to enable students 

 to distinguish the eggs of one species from those of another is 

 futile. So far as I know, no attempt has been made to establish 

 a m?.an description of any sjiecies — that is, a description based 

 on statistics. 



While we may observe sets of A and B alike, C will be different : 

 hence, descriptions based on statistical methods would help one 

 to arrive at a normal type — that is to say, a type which occurs 

 most frequently in our observations. One hundred sets described 

 and measured may prove to have 35 approaching A, 55 of the 

 B type, and 10 of the C type. A and B being similar, the mean 

 or normal type would be derived from them. 



It is manifest that a more expeditious method of describing in 

 detail is needed, especially for the use of future generations, who 

 will have, perhaps, nothing but descriptions handed down to 

 them to work upon, as rare types will not be available for students. 

 Great advantages would accrue by the detailed description of 

 rare species, such as Atrichia rufescens, Pfilorhis paradisea, &c. 

 This system would be distinctly advantageous in describing 

 type sets, as the salient features of each egg could be treated 

 minutely. Much verbiage will scarcely succeed in conveying a 

 definite idea unless accompanied by a concrete guide. The 

 oologist describing an egg thinks his description perfect ; but the 

 student who has to educe a mental picture from the describer's 

 words is apt to strain his imagination, and is at a great dis- 

 advantage compared witli the describcr. who has the actual 

 specimen in view. 



