300 BRITISH LEPIDOPTERA. 



variation has been traced by Chapman, who shows that there is, 

 in all Fumeids, a large basal joint, then a largish globular joint, 

 before we come to the flagellar joints carrying the pectinations ; 

 occasionally (and this seems to be an individual and not a specific 

 character) there intervenes between the 2nd joint and the normal 1st 

 joint of the clavola (or flagellum) a very short joint without pectina- 

 tions. In one instance, joints 3 and 4 (i.e., joints 1 and 2 of clavola) 

 are fused together, making a long joint with four pectinations. The 

 last flagellar joint also is usually a simple one without pectinations, 

 but occasionally this is very small, and at times wanting, and then the 

 last joint is one with two short pectinations. Based on the number 

 of antennal joints we get the following division : 



(1) 20 joints or more: Reticulatella (22), comitella (20), norvegica (22), crassior- 

 ella [24 (one race), 22 (second race)], affinis (24:),subflavella (24), edwardsella (20). 



(2) Fewer than 20 joints: Mitfordella (17-18), hibernicella (19), casta ab. 

 minor (18), casta (type) (18), casta var. intermediella (19-20), bowerella (16-17), 

 scotica (19-20), gcrmanica (19-20). 



Chapman observes that the length of the antennal joints are 

 fairly, and the length of the pectinations very, uniform. The great mass 

 of specimens have pectinations between -49mm. and -51mm., whilst the 

 length of the antennal joints is less uniform — four crassiorella vary 

 from "lOInim. -•195mm. ; affinis, subflavella and casta from -200-"208 

 (except one of the latter -210) ; intermediella runs from -21 6- -227, 

 bowerella *240, germanica -182-'194. There is too little material 

 perhaps for generalisation, but the character may be of specific value in 

 bou-erella, and as a general rule it may be stated that pectinations are 

 of uniform length throughout the genus, but that the length of 

 antennal joints varies directly with the size of the insect and inversely 

 as the number of joints. 



The character on which most reliance has hitherto been placed in 

 distinguishing species has been the form and size of their wings and 

 even their colour. As we have already noticed, reticulatella and comi- 

 tella are distinguishable by possessing definite wing markings, but the 

 remaining species possess no reliable characters in this direction. 

 Colour is certainly of no real value, all the species go brownish with 

 age and many are of varying colour when fresh. The form of wing 

 is certainly variable within the limits of one species, but is not so easy 

 to define as constant between different species. In the British forms 

 that have been divided into various species the apparent form of the 

 wing is possibly much more dominated by various styles of setting 

 than by any inherent character. All the species have a certain 

 amount of natural curvature in the wings and also a certain amount of 

 pleating or folding at nervures, and the extent to which these are 

 straightened out by pressure and flat setting, or exaggerated by droop- 

 ing, makes an immense difference in the apparent shape of the wings. 

 There appears also to be considerable variation in size within the 

 limits of a species. Some species, however, are distinctly larger than 

 others, the largest being crassiorella. The apterous females also 

 present certain characters, but Chapman, to whose work we are again 

 indebted for the material in this part of our book, states that he has 

 so far been able to differentiate only two forms associated with known 

 species, these are — crassiorella and our common British casta. He 

 separates that of crassiorella by its larger size, dark vinous colour, and 



