SCIEXCE-GOSSIP. 



the banc-formulae ooooo, 00300, 00345, and so on. 

 The form ." bellula may be subdivided in the same 

 way, but reduction in an already pale yellow is not 

 very noticeable. In all the nemoralis considered, 

 : = io per cent., libellula = 20, 1; '..'.'.1 = 2S. 

 42. Some of the more noticeable points are 

 the preponderance of (22 per cent.) from 



Dormington Quarries, and the high percentage of 

 libellula at lloreton (31), and in the De Boinville 

 collection (36). The very low (7-5 per cent.) 

 number for rubella, from Moreton, is compensated 

 for by the high percentage (56) of mista, so that 

 the tea! mista and rubella is not very far from an 

 average. For further details see Table I. 



In Tachea hortensis the lack of variation in this 



but the great majority are in a border class of 

 their own, and to make a fresh group for them 

 would only double the difficulty by having two 

 boundaries instead of one. The colour varies 

 considerably on different parts of the same shell. 

 Reduction of ground-colour here too appears to 

 follow strong banding ; the fewer the bands the 

 deeper the ground-colour in lutea, and hence a 

 heavily-banded lutea with a bright-yellow ground 

 is uncommon ; in true alba the paling of the ground- 

 colour is due to something else. 



(d) var. arenicola, Macgill (S per cent.), with trans- 

 parent bands. Of this there are two fairly distinct 

 forms, the one with an opaque, the other with a 

 translucent ground ; the latter are, of course, the 



Table I , showing the percentage 



OCCURRENCE 



OF THE TWO SPECIES OF 



TACHEA, OF 



THEIR 



VARIOUS COLOUR 



FORMS, AND OF UNICOLOROUS AND 



BANDED SPECIMENS. 











Locality. 









f~ 



( 



) Dcrinington 







(4) Raihvav 





•» 







Quarry. Dec., 



(2>De 



More- 



Bank, 



(5) Broomy 









and 



Boinville. 



ton, July. 



Hunderton, 



Hill, 



4 and 5 







Jan.. 



X&43, circa. 



1S06. 



April, 1S95. 



April, 1S95. 



combined. 



Average. 



No. of specimens 



. . 



737 



54 



2l6 



30S 



524 





T. nemoralis . . 



. . 90-9 



74'5 



833 



44'4 



27-6 



36 



54-68 



- 



. . 222 



113 



5'i 



62 



71 



4-135 



I0-3S 



unicolor 



. . IOO 



17" 



100 



667 



100 



S3-35 



76-SS 



fasciated 







82-3 







333 







16-65 



23-12 



var. libellula . . 



.. i 3 -0 



36-1 



309 



14-6 



!5'3 



I4-95 



20'10 



unicolor 



. 



7-8 



















5'2 



fasciated 



.. 81S 



92-2 



100 



100 



100 



100 



948 



var. rubella . . 



.. 34- 



3S2 



7'5 



33'3 



25'9 



296 



27-90 



unicolor 



107 



10-9 



2-9 



22O 



136 



i8-i 



12-14 



fasciated 



-• 89-3 



89-1 



97-1 



77'4 



86-4 



81-9 



87-86 



var mista 



296 



144 • 



56-4 



469 



517 



493 



41-62 



var. mista and rubella 64-2 



526 



640 



79-2 



77'S 



78-4 



69-52 



unicolor 



•• 3"l 



80 



o-35 



9-2 



4'5 



6-85 



5-03 



fasciated 



• • 96-9 



920 



9965 



908 



95'5 



93-15 



94-97 



Total unicolor 





87 



5'3 



115 



io'O 



11-05 



12-90 



Total fasciated 



.. 716 



913 



94 7 



88-5 



S94 



8S-95 



87-09 



rttnsii .. 



.. 91 



255 



167 



550 



724 



64 



45-32 



var. alba unicolor 







re 



11 



— 



i'4 



0-7 



0-82 



var. lutea unicolor 



.. 28 6 



154 



855 



433 



33-G 



38-45 



40-2S 



fasciata 



-'•4 



505 



Hi 



54 2 



619 



5S-Q5 



49-82 



var. a»i 







30- S 



2-2 



2'5 



3'i 



2-8 



772 



var. fu:u unicolor 



— 



1-6 



— 





— 



— 



032 



Total unicolor 



. . 28 



18-6 





433 



35'° 



39-I5 



42-44 



Total fasciated 



.. 714 



814 



13 3 



5 r ---7 



650 



60S5 



57'5'' 



neighbourhood is still more marked than it is in T. 

 Htmcrralii. Here again I can distinguish four main 

 types 



per cent.) : one must 

 reckoning pale lutea under this head, though 

 the distinction is really not one of any very great 

 importance 



ar. lutea unieclor (41 per cent): varies a great 

 deal in intensity of colouring 



\r fauiata (jo per ccr ■ he same 



enke'i fauiata . I apply this term to white or 

 specimens posscsv: I have 



quite failed to distinguish alha from h 



-.cur (more commonly in alba) 

 fairly frequently, and a 



arenkola form of thin translucent lutea or alba. A 

 tinge of brown in the bands— which always seem 

 to be pigmented transparencies — places the speci. 

 men in the fasciata group. 



A fifth form has occurred— fiisca, Poiret— to the 



extent of four specimens in the county : three in the 



inville collection, and one from lireniton 



rid 7NJ, which last is not reckoned in this 



paper. The explanation of the lack of variation 



probably lies in something beyond the geological 



tony, as a much more extensive range of 



■ forms occurs in other apparently equally 



mon- dl iricts. 



illection is rcmarkabl'- foi a 

 (31) i 11 ■" to, " Uli b.1 



