Fossorial ITymenoptera of North America. 51 



head or vertex, varies as that region is elevated and very convex, 

 or flattened and broad longitudinally. They are all placed nearer 

 together in a nearly equilateral triangle, not differing so much as 

 in the ('rethrows. In C. deserta when the vertex is convex, they 

 are situated upon the top, but in other species of Cerceris where the 

 vertex is more flattened the hind edge rises above them a little. In 

 Philanthus when the vertex is rounded above, they are very near the 

 top. In Pemphredoninse, they are above the vertex. 



The eyes in Cerceris are oval, whole; very slightly indented by the 

 epicranium in Philanthus, when they approach one another on the 

 vertex in Pemphredon, &c, they are rather smaller and within very 

 straight and diverge towards the vertex. 



The front is flat and square in Cerceris, &c. Square with the cor- 

 ners rounded in Philanthus, and also full slightly convex. In Pem- 

 phredon, &c., hollowed, and the sides converging in front, and in 

 Mettinus* full again, much like Philanthus; in Gorites imperceptibly 

 depressed, triangular. 



The insertion of the antennae varies much in situation. In Cerceris, 

 they are placed above the middle of the front, (especially in C. deserta') 

 in Philanthus near the middle and far apart, in the remaining genera 

 below the middle (especially in Pemphredon very near the margin, as 

 in the Crabrones, where they are placed very near together. In Cer- 

 ceris and Philanthus the second joint is globular and very short. In 

 Pemphredon, long and slender. The remaining joints thicken con- 

 siderably beyond the middle, especially in Philanthus. In Pem- 

 phredon they are smaller, shorter and of more uniform thickness 

 throughout. 



The history of the clypeus is very interesting, and its changes of 

 form in the different genera are very great, as in the Vespidae. In 

 the Cercerida; it varies greatly, while more constant in Crabronidae. 



In (C. deserta) the clypeus is trapezoidal, broadest and nearly 

 straight in front and flattened. In other species of Cerceris it is 

 shorter, much raised and nearly square, (I speak of the " middle lobe" 

 of authors, for I much doubt whether the " lateral lobes" belong properly 

 to the clypeus, but are not rather the sub-mandibular pieces of the man- 

 dible and belong to the epicranial pieces, though indeed we see no 

 distinct suture between the two lobes in many genera, yet in C. de- 

 srrta there is a distinct suture, and as the presence of these lobes is 

 one of the best characters of the family, giving the breadth to the 



* These genera of Xyssonidse are introduced for the sake of comparison. 



