1900 ] 77 



and well defined by a suture. Below the labrum the labium appears at first without 

 a central suture, which comes lower, before it disappears under the maxillae. These 

 are wide basally, long, extending nearly to end of wings. The second legs extend 

 exactly to end of wings. The third legs, appearing below end of maxillae, extend 

 just beyond wings. The antennae are longer than the maxillae, but do not quite 

 reach the end of wings. The first legs are much shorter, to three-fourths of the 

 wings. Between them and maxillae a portion of let femur (?) appears. Above the 

 top of second leg is a well developed " eye collar " (end of maxillary palpus). The 

 tips of the wings and appendages extend a little beyond 4th abdominal segment and 

 hang over the 5th or intersegmental membrane, according to attitude of pupa. Hind- 

 wing well seen to middle of 3rd segment, but still shows a minute border round 

 to apex. 



This pupa is very interesting, as being one of active habits, 

 although as fully obtect in structure as any other Pyrale. The 

 modifications of structure that enable it to move so freely are in no 

 respects remnants of structures that are used to produce such move- 

 ments in pupcB incompletce, but are developments from an obtect form. 

 The pupa is in a silken tube of several times its own length, which is 

 usually under a stone or amongst lichens and herbage, and has one 

 extremity near the surface and the rest of its length more or less 

 buried. The pupa is able to travel easily along this tube, and es- 

 pecially can very rapidly run backwards to the remotest extremity of 

 the tube. I should imagine it comes to the front of its habitation to 

 obtain the benefit of sun heat, retreating during the night, or on any 

 indication of danger. It does not leave the tube for emergence. The 

 retreat is made rapidly by aid of the great spatular cremastral arma- 

 ture, each straightening of the body followed by curling forwards of 

 the hinder extremity, carrying the pupa backwards nearly one-fourth 

 of its length. The advance is made more slowl}^, and seems to be 

 effected by the dorsal and lateral spines, which are not homologous 

 with the rows of spines so common in piipce incompletce, but are modi- 

 fications of the ordinary tubercles which we are familiar with in larvae. 

 It is curious that the supra-spiracular tubercles are present, but only 

 one pair of the trapezoidal, apparently the anterior, but this is not 

 easy to feel positive about, whichever they are (I or IT) no trace 

 whatever exists of the other pair. It is also noteworthy that on the 

 posterior segments (5-9) the general surface has attained the first 

 stage in the line of development, which appears to have been passed 

 through in Incompleted^ in attaining the rows of doi'sal spines, viz., the 

 general surface has roughened so that each cutaneous cell instead of 

 being smooth carries a fine point. It is easy to imagine that had the 

 tubercles not developed so as to meet the requirements, an armature 



