314 Mr. Poulton's further notes upon the 



anal tubercles. The lateral margins of the anal flap 

 are invisible, except upon careful observation. A few 

 scattered hairs are still present, and are rather more 

 abundant anteriorly and posteriorly. The ground colour 

 varies from red-brown to dull yellow ; in all cases 

 mottled with lighter tints. The anterior white ring of 

 earlier stages is present as two short transverse pale 

 yellow lines, one on each side of the dorsal middle line 

 in the anterior part of the 2nd abdominal segment. An 

 inconspicuous lightish cloud sometimes extends back- 

 wards from these marks for a short distance. The 

 transverse markings are sometimes covered by the 

 reflected posterior part of the 1st abdominal segment in 

 the protective attitude. There are no traces of the next 

 two bands, but the light colours on the side of the swollen 

 4th and 5th abdominal segments are the remains of the 

 fourth band, which became diffuse and cloudy in the 

 second stage. The 4th abdominal segment is not 

 covered with the light cloud to such an extent as the 

 5th. There is a tendency towards a longitudinal 

 arrangement of the lighter markings, especially in the 

 more cylindrical parts of the larva. The under side of 

 the head and 1st and 2nd thoracic segments is much 

 lighter, and of a dull yellow colour. This of course 

 becomes the upper side in the protective attitude (see 

 fig. 17, Plate VII.). The spiracles are very indistinct at 

 all stages. The segments behind the fifth abdominal 

 are darker in colour, generally showing a combination 

 of very dark grey and rich brown. The ventral surface 

 of the first four abdominal segments is rich brown with 

 creamy mottlings longitudinally arranged, the colours 

 being continued on to the under side of the transverse 

 ridge bearing the 3rd pair of true legs. The upper 

 surface of this ridge is very dark. In many larva3 the 

 brown is replaced by greyish tints. The protective 

 attitude of Stage III. is remarkable for its excessive 

 irregularity in the vertical plane. Occasionally there is 

 a deviation to one side, especially in the bend at the 

 3rd thoracic segment ; but this is exceptional, and the 

 efficiency of the attitude is not, as a rule, due to a 

 spiral or bilaterally asymmetrical position, as in the 

 case of the Ej)hyri(he. Nevertheless, the larva at rest 

 is wholly unlike a Geometer in the usual position of 

 resemblance to a twig, although this attitude is assumed 

 during later stages. During the maintenance of the 



