24 
thirty he thinks he may be occasionally wrong; at five-and-forty he 
knows he is occasionally wrong; at fifty-five he has learned his 
ignorance, and gives his opinions to the world in fear and trembling. 
Of course, there are butterfly-catchers whose age reaches the know- 
ledge of ignorance standard, but whose sagacity includes them with 
those aged five-and-twenty. 
This is an excursus, and I apologise. I would only impress on 
my hearers that species are not quite entities; that they are not 
absolutely defined by exact limits, but yet that they are sufficiently 
definable as a rule, that, with a little latitude, a name may be applied 
to them which will convey to the initiated a more or less clear con- 
ception of what the user means thereby. 
But if species are not absolute entities, what is to be said of genera 2 
At the outset it must be conceded that although genera are not 
imaginary creations, yet they are by no means clearly defined, 
abruptly separated from each other, nor well marked, except by 
accident, but merge into each other by slow and easy gradations. 
That genera do merge slowly into each other is no excuse for uniting 
them into one genus, and hence the necessity for defining a genus by 
characters obtained from all the stages of existence of its species, for 
it is only by doing this that the true position of what we may term 
intermediate species can be obtained. ‘To take a single character, 
or even characters, common to adjacent groups, so that the various 
species may be enumerated with equal plausibility under two con- 
secutive groups, and then infer that the groups cannot be upheld, 
and forthwith proceed to annihilate the genera by blending them into 
one so-called genus, isa plan much to be deprecated, and 1s, in reality, 
an attempt to fall back, to a certain degree, upon the view that 
species alone are to be recognised i in the organic world. 
In the early days of zoological science it was to be expected that 
general and obvious characters would be the first to be taken and 
used. Hence the massive genera of the older entomologists, to 
whom all butterflies were Papilio, all Lepidoptera with pectinated 
antennze were Bombyx, all Plumes were A/uci/a, and so on. It took 
almost a century of slow progress before our entomological ancestors— 
Stephens, Doubleday, Westwood, and others——w eal out the details 
of the minor characters separating closely allied groups, which gave 
us a more or less natural system of characters by which we could 
recognise and separate allied groups. For the last fifty years, how- 
ever, our advanced lepidopterists have been engaged (1) in the 
working out of life-histories, and (2) more recently in discovering the 
phylogenetic characters presented by the creatures they study ; and 
during this time no capable scientist has thought fit to generalise on 
the material thus collected, and so give us an adv anced and natural 
scheme of detailed classification ; nor have our book-makers, who 
have compiled books from which collectors could name their 
captures, attempted to set forth the alliances of the various insects 
described, according to the latest discoveries. Not only has this 
