412 MR. E. SAUKDEES ON THE 'lONGTIES OP THE 



apparatus lies folded when at rest, and I shall here refer to it as 

 the oral groove. By gently pulling the apex of the tongue, the 

 whole apparatus may be gradually unfolded, and it will then be 

 seeu that there is a membrane investing the entire base of the 

 apparatus and completely covering the oral groove ; this is best 

 seen when the tongue is fully extended. On dissecting the head 

 it appears that this membrane is attached to the back of the 

 clypeus on each side, whence it arches downwards and forms 

 what has usually been termed the hypopharynx, between which 

 and the epipharynx is a very small opening, and it is through 

 this opening only that there is communication with the gula, 

 except through the actual lingua itself. The cibarial apparatus 

 may, in fact, be compared to a funnel-shaped bag strengthened 

 by various sclerites above and below (which also aid in its folding 

 up), and prolonged at its apex into the semitubular mentum and 

 lingua. The arrangement of the various sclerites is as follows. 

 I shall commence with those on the underside : — 



On the anterior edge {i. e. that lying furthest from the back of 

 the head) of each of the truucatures that bound the oral groove 

 is an emargination, into which articulates an elongate joint, these 

 joints are called the cardines ; each cardo is considerably widened 

 at its apical end, and produced into two unequal processes. On 

 these widened portions swing the lora, and from them depend 

 the maccillce, each of which consists of a sheath-like basal portion 

 and a blade-like apex, between which on the external side the 

 maxillary palpus is inserted. The lora are two narrow joints 

 united so as to form a ^-shaped body, from the angle of which is 

 suspended the mentum, &c. ; these can swing over on their feet 

 which rest on the apices of the cardines, and by this motion the 

 tongue can be projected for twice the length of the ^. They 

 vary very much in their length, and in some of the genera of the 

 short-tongued bees are scarcely developed ; but in only four 

 British genera are they actually wanting, viz. in Salictus and 

 SpJiecodes, and in Dufourea and Bophites, in which there appear 

 to be no traces of them, and, in fact, there appears to be nothing 

 to correspond to them in their ofiB.ce of lengthening the tongue ; 

 the membrane covering the space between the cardines, which in 

 the other genera is simple, is here chitinized towards its centre 

 into two straps, which' are united to the submentum at their 

 apices. Prom the angle of the lora hangs the submentum, which 

 varies considerably in shape; although always more or less 



