2 LEPIDOPTERA HETEROCERA 
Pupa with first four segments of abdomen fixed; sometimes with rather prominent scattered hairs 
(group 8); usually in slight cocoon amongst refuse or on food-plant. 
* 
* —* 
The conception of the family Gelechiadae is apparently due to Stainton, but (as noted in my 
remarks on the Oecophoridae) he did not discriminate properly between these two families, their limits 
being defined by myself in 1883. Subsequent discoveries of material have shown that (as indicated in 
the family characters above) there is really not a single character to which an exception does not occur; 
as knowledge widens, all boundaries become less absolute. I am however still satised that the — 
Gelechiadae and Oecophoridae do represent two natural associations or lines of development, arising from 
a common base and not intermixing at any other point, but in order to distinguish the two in practice it 
is necessary to consider the whole of the characters in conjunction. and not to argue from single points, 
There are in fact many points of distinction between the two families, and the failure of one can always 
be corrected by the inspection of the rest. In A4momoxena a character (common to both families) fails 
which might have been thought (and probably really is) difficult of modification, and to which no other 
exception is known amongst the whole of the 7oo genera of the two families, yet this is in a highly 
specialised genus far up the line of ascent ; as is easily seen from the rest of its characters. Great caution 
is therefore advisable in accepting the statements of those whose limited experience recommends faith 
in any particular characters whatsoever. 
I have divided the family into nine groups ik I consider undoubtedly natural, but I have not 
ventured to attach names to these groups, because they are not properly definable; there are in all cases - 
structural characters which serve as guides to their identification, but there are so many exceptions that. 
it would only be misleading to draw up definitions. When familiar with these groups, however, it is - 
generally possible to say at once of a new genus to which group it should be referred, and thus they are - 
really serviceable as mental conceptions, in view of the large size of the family, and contribute. 
materially to its comprehension. 
The family as a whole must be regarded as a rather modern development, which has MAS 
.in adapting itself to a great variety of situations. In New Zealand it is very scantily represented, 
obviously only by three or four types that have found their way there in recent times. Elsewhere it is 
generally abundant (not in oceanic islands), but with different types predominant in different regions; 
thus in Europe and North America the Gelechia type predominates, in India the Lecifhocera type, in 
South America the Compsolechia type, in Australia the Protolechia type; these groups are therefore more 
recent in origin than the period since which these main regions have been disconnected. These 
distinctions are very pronounced ; but owing to the obscure appearance and secretive habits of many of 
these insects, there are undoubtedly a large proportion still awaiting discovery. 
In the present revision I have classified 391 genera and 3542 species. 
KEY TO THE GENERA 
1. Forewings with naked glandular spot beneath costa towards base .. 64. Genus CrisTOTHYRIS, Zeller. 
Foreninm Wut Suh spot. 0.500 osos ie ow La SR oe LL MR NR RE a. 
2. Forewings with 7 and 8 separate... . . . . . . . 33. Genus ANOMOXENA, Meyrick. 
Forewings with 7 and 8 stalked or coincident à ERROR NC o atia ap ER ^ 
3. Forewings wilh 7 lo lermen or apex, or 7 (or 8 or 9) dies NS oL) EDIT ER DN pru 4- 
Feresingr ,U/À 2 do swa, 7-Q AIQM 245 5 797. odo. n. 9 To Udo ow 4 Uo a sio UB. 
i, FoPewios AE D do cota. 2 8 o 0o aAn cS T-u V RpTN EE RU es 3. 
Forenage quib .6 abes or HR 2 2 9 65s P DAS v cueva ER Qe Yr 6. 
