Woodriiffe-Peacock : Thrush Stories , etc. 173 



This is quickly done by adding a colon after the figures that 

 want less than one type-band increase, and a point after those 

 that require diminishing to a like extent. By doubling or 

 quadrupling the figures, a perfectly exact formula for any shell 

 may be obtained. Difficulties of several kinds are met with by 

 doing this, both in the field, and in clearness of recording, so 

 in practice such formulae are unworkable. 



Three shells from a high hedge bank on sandy glacial gravel 

 will illustrate at once this method and its flexibility. Here are 

 their formulae :— ii2225(425), li.2i.3:45i6, and l:i.2:2.1:6.5i.5:. 

 There is nothing in anyway unusual about them. In the first 

 shell the upper band was missing, and the lower ones confluent. 

 Small figures always imply that the band and interspace, 

 or the bands arxl interspaces, as the case may be, are absent^ 

 but that the space covered by them may be thus approximately 

 accounted for. In the second shell, as their formula records 

 them, the two first interspaces are too large, and the third band 

 too small. The third shell is a more difficult task to take off 

 correctly ; the fourth band alone is typical, all the other bands 

 and interspaces require diacritical marks of increase or decrease. 

 Along with these specimens two other shells were brought home 

 from the Oxford clay of a dyke side. A lihellula (Risso) + conica 

 (Pascal) which read 1.21.22:52:4.5; and a rubella (Moq.) + 

 compressa (Terver.) which read li2225425, i.e., with a simple 

 band formula. 



With this fairly expeditious method — when it is fully mas- 

 tered — the bands and interspaces of H. nemovalis from varying 

 localities and soils can be formulated sufficiently accurately 

 for practical scientific results. The law of their relationship 

 to their environment, and of the frequency of the destruction 

 of all forms can be worked out. Still more important, the 

 evolutionary law ' of the correlation of parts or characters ' 

 can be discovered so far as the bands and interspaces are in- 

 fluenced by it. The sheets I use for recording purposes are 

 three inches wide by three and three-quarters long. I make 

 the most exact notes of locality, soil, water, etc., for every- 

 thing seems to influence the banding and interspacing, i.e., the 

 destruction of this species by thrushes. I keep the notes under 

 soils, arranged in the order of their colour in the first place, then 

 their varieties of form, and finally by the number of bands, 

 ignoring the interspaces. 



The commonest shell found at thrush stones is lihellula 



iqog May i - 



