172 Woodruffe- Peacock : Thrush Stones, etc. 



similar range on the bank of another stream. Personally, I 

 have never taken this form anywhere except by flowing water. 



Again, when the forms of banding of various soils and locali- 

 ties are brought together and are compared, the general band- 

 ing law of the species is clearly seen. For H. nemoralis, it is 

 most usefully stated in a formula, as li2234425, which notes 

 both the banding and inter-spacing widths. In these figures, 

 reading from left to right, the first, third, fifth, seventh and 

 ninth indicate the bands, and their general normal width ; 

 while the second, fourth, sixth, and eighth figures indicate the 

 inter-spaces and their general normal width. In other words, 

 the space between the upper side of the first band, and the 

 lower side of the fifth band is divided into 24 imaginary band- 

 spaces of equal width. As 24 is a number that can be divided 

 by many other numbers without leaving a fraction, no other 

 number about its size could be found equally useful. Even 

 when the shells are not typical, such as the varieties major 

 (Fer.), minor (Mog.), compressa (Terver.), or conica (Pascal), 

 the formula applies. 



In practical band and interspace recording in the field and 

 study, I find it impossible to take off the banding formula of 

 a shell in one long line as it is printed here. I used to write 

 the typical formula down with the bands at wide distances, and 

 below an imaginary line, the interspaces between them. Then 

 I took the shell which I was about to record, and studied it 

 to see which had the greater width, the fifth band or third 

 interspace, and from this drew a criterion for the data of the 

 shell ; writing its formula in the same way below that of the 

 typical shell formula. I know it so well now, I have no occa- 

 sion to write the type formula ; but that of the shell I still 

 write in the same way, and strongly advise all analysts to do 

 so. The eye is much helped by two lines of figures — one for 

 the five bands, and one for the four interspaces, though they 

 cannot be printed in that way conveniently. In reading off 

 the printed formulae, too, much assistance is gained by noting 

 the central figure specially, which is always the third band record 

 for every shell. 



Specimens that are accurately represented by the type or 

 any other simple formula are comparatively rare. So a prac- 

 tical method has to be discovered to show at a glance that the 

 recorded bands and interspaces are wider or narrower than the 

 width of the one twenty-fourth of the whole banding area. 



Naturalist, 



\ 



