218 ON SYSTEMATIC ZOOLOGY. 
thrown great uncertainty on his circles of Ametabola 
and Crustacea, but alters the position of every group in 
the diagram. If, on the other hand, we are to believe 
that the positions of these groups in the annulose dia- 
gram is correct, then that of the Mandibulata* must be 
reconstructed ; for the Thysanuriform type cannot be at 
once both typical and aberrant ; nor can the Chilognathi~ 
form Ametabola be aberrant, and the Chilognathiform 
Coleoptera typical. We are convinced, therefore, that 
this talented author had not ascertained the fact, that 
the denomination of a group is always definite ; that is 
to say, it is either always typical or always aberrant : 
he justly supposes that the contents of one natural 
group represent the contents of another natural group; 
but he did not perceive that one of the consequences of 
this fact was, that the divisions which were typical in one 
would be typical in another ; for if otherwise, the parallel 
between the two would fail. This oversight, in fact, 
has not only proved the artificial nature of the ame- 
tabolous circle, but has been no small source of em- 
barrassment to the attempt of arranging the order 
Lepidoptera in conformity therewith. Every entomo- 
logist must perceive that the two typical (or external) 
divisions of the diurnal butterflies, as Shrank and the 
authors of the Vienna Catalogue long ago intimated, are 
represented by the genera Mymphales and Papilio; the 
first being the Chilopodiform stirps of Dr. Horsfield, 
and the latter the Chilognathiform: but if the series of 
the Ametabola is to be adhered to, as given in Hor. Ent. 
p- 390., then this theory must be altogether abandoned ; 
the genus Morpho, as corresponding to the Thysanura, 
must be a typical group ; that of Nymphales, the other ; 
while that of Papilio becomes aberrant, leading directly 
out of the circle! It is clear, therefore, that if, as many 
passages in his work indicate, our learned author enter- 
tained a suspicion that the rank of his groups was de- 
finite, he did not believe that this property was univer= 
* Hor. Ent. p. 489. 
