NOTES ON THE FUMEIDS. 61 



and carries two very short pectinations, which are longer on the 4th 

 and reach full length on the 5th or 7th, dwindhng again gradually to 

 the penultimate joint, where they are short, the last joint being simple. 

 In several instances, apparently of individual and not of specific 

 variation, however, there is intercalated between the 2nd and 3rd 

 joints as above described, a joint that is very short and disc-like, with- 

 out pectinations ; in others three and four as above described are fused 

 together into one long joint with four pectinate processes. Similarly, at 

 the extremity of the antennas, the last simple joint may be of various 

 lengths or may be wanting, the then last joint being pectinated, or 

 sometimes it might be more correct to say that the last normally 

 simple joint possessed pectinations. 



The examination of the antennas is unfortunately handicapped by 

 two circumstances. In the first place the antennne are very apt to be 

 damaged by mould, mites, Psoci, or other destructive agencies to a degree 

 much beyond anything in most lepidopterous insects. In the second, 

 they can only be confidently described when removed from the specimen 

 and mounted in balsam or otherwise, and this is quite inadmissible in 

 regard to many specimens, in fact, in regard to any not one's own 

 property, unless by special permission. Yet, without this, one cannot 

 certainly count the number of joints, as the first pectination varies a 

 good deal in length and in the extent to which it is visible bej'ond the 

 thick scaling of the basal joints and of the face. As a rule, I think, 

 descriptions stating the number of antennal joints mean what they 

 profess, viz., actually all the antennal joints, and may or may not be 

 accurate owing to the difficulties I have mentioned, but may be taken 

 as being to a high degree of probability within one of the truth. In 

 other cases, however, I think descriptions merely cite the number of 

 visible joints, that is, of pectinated joints visible in a good specimen. 

 This is the case, for example, unless I am much mistaken, in the dis- 

 tinction drawn by Hofmann (Beiiiner Ent. Ztschft., iv., p. 32) between 

 the antennas of nitidella and ajfinis, Avhere he gives the former sixteen 

 joints and the latter 21, when truly they have eighteen and 23 (or 

 more properly 21). 



I have carefully measured the length of the antennal joints and of 

 the pectinations in a number of specimens. The uniformity of these 

 throughout the genus is quite beyond what I expected. There are one 

 or two anomalies that may be of some definite meaning. The great 

 mass of specimens have pectinations of a length between -49 and -51 

 of a mm., and the length of a joint varies from -183 to -210. The 

 exceptions in the pectinations are an a finis, which has them "SCmm., 

 and an intermcdiella, preserved in balsam four years ago and probably 

 placed under the cover glass shortly after emergence and subjected to 

 pressure. This may account for the otherwise anomalous figure of 

 •67 — the joints in this specimen are also long, '216 ; or it may be 

 that these are the correct measurements and that those taken from dry 

 specimens are in error owing to contraction. However this may be, it 

 remains that fifteen antennae, belonging to M. crassiorella, M. sub- 

 fiavella, M. edivardsella, M. affinis, F. intermediella, F. casta, F. 

 germanica, and F. howerella, have pectinations, of which the largest do 

 not vary beyond from •49mm. to -SSmm. in length, a dift'erence well 

 within errors of measurement. The length of the antennal joints is 

 less uniform. Four M. crassiorella vary from •164mm. to •195mm. 



