304 Mr. Poiilton's//w^/i6'/' notes iip<ni tlte 



vessel. This was often seen in S. ligustri and S. ocellatus, 

 by the use of a lens magnifying fifty diameters. The 

 movement always took place, and was a source of con- 

 siderable difficulty when I tried to obtain an outline 

 drawing of the horn by means of the camera lucida. 

 It seems that this movement (which is in the vertical 

 plane) is less in amount before ecdysis. 



10. Instances of the probable passing backwards of 

 CHARACTERS IN THE ONTOGENY. — In the abovc-described 

 ontogenies there were certain cases in which we seem to 

 witness the actual passage of characters backwards into 

 an earlier stage than that in which it had previously 

 appeared for the first time. Thus in S. ocellatus a very 

 small proportion of the larvae in the first stage possessed 

 the specialised head of the genus, the others having a 

 rounded head. The gradual acquisition of this character 

 by the first stage is to be expected, for this period has 

 already acquired everything else that is distinctive of 

 the second or even later stages. Again, in S. ligustri 

 the purple borders to the obKque stripes appear, as a 

 rule, in the third stage, but the time at which they 

 appear and the extent to which they develop are very 

 variable, and sometimes they are not present at all. 

 The instances in which the borders did not appear until 

 the fourth stage were, however, rare in my experience, 

 but they seem to have been universal in the case of 

 other observers. Here, then, is a character which has 

 nearly, but not quite, established itself in the third 

 stage. So also with the forked termination of the 

 caudal horn in S. ocellatus and S. ligustri. This 

 character is fading out of the advanced stages of these 

 two ontogenies. In the first stages it is always present, 

 but later the structure is very irregular in the degree to 

 which it is developed, and it is only present in a certain 

 proportion of the larvae, that proportion becoming 

 smaller as the stages advance. 



11. The importance of maintaining the continuity of 

 SERIAL MARKINGS. — I have already pointed out (last year) 

 that a slight suggestion may produce the effect of a 

 continued series when the marking has been repeated 

 sufficiently often. If the series be of protective value 

 such suggested continuity may be of great importance 

 to the species. Conversely a very decided break in an 



