OF MANDIBULATA. 4%3 



Linnaeus which have apod larvae (the Larva: subvermi- 

 formes of Latreille) to form one group, it seems not im- 

 possible to refer them all to one or other of the following 

 subdivisions. 1. Anthophila Lat. 2. Rapacia of Lamarck, 

 excluding his second division. 3. Pupivora Latr., with 

 such modifications as will make the group consist only of 

 those insects \vhich have a plurivalve fissile appendage to 

 the abdomen. 4. Iieterogyna Latr. : and o. Tubulifera Latr., 

 which last appear to approach in habit, organs of man- 

 ducation, and structure of wings, to some brilliant In- 

 dian bees, allied to Latreille's parasitical family of No- 

 mada. 



It is undoubtedly through the medium of certain Pa- 

 pivoraot Latreille that we ought to quit the Hi/menoptera. 

 If we look to external structure, the genus Aulacus of 

 Jurine will probably be fixed upon ; and if we resort to 

 theory, and ask which of the true Hynmioptera have 

 phyllophagous larvae presenting vestiges of feet, it may be 

 answered that certain species of the Linncean genus Gy~ 

 nips correspond with this description. The truth may 

 possibly be found hereafter to lie between these two sup- 

 posed means of arriving at the genus Tenthredo of Lin- 

 naeus. Sufficiently certain, however, it is, that the Ura- 

 cerata of Latreille, composing the genus Sirexof Linnaeus, 

 come nearer than Tenthredo to the Hymenoptera. The 

 abdominal appendage of Sirex is constructed on the plan 

 of that of the Ichneumonlda, and the larva has only six 

 feet; whereas in the true Tenthredines the oviduct is com- 

 posed of four pieces, of which the two internal are serrated 

 and sheathed by the two external, and the larva at last has 

 from eighteen to twenty-two feet, of which all after the 

 first six are membranaceous. 



