166 PROCEEDINGS ENTOMOLOGICAL SOCIETY 



. 



larva is just as liable to specialize in view of changed conditions 

 as the adult one is." Dr. Chapman is of the opinion that where 

 the sapfeeding habit has actually been acquired in two or three 

 instars only, it will not continue to encroach on later instars. 

 His idea that "Any larval instar may undergo changes without neces- 

 sarily involving any other instar" may be true generally speak- 

 ing, but in the Gracilariidae, in as far as the sapfeeding habit is 

 concerned, it does not seem to hold, if any significance is to be at- 

 tached to the existence of intermediate forms. Thus, Acrocercops 

 strigifi.nitella, in its third instar has the head and mouthparts prac- 

 tically normal, but the body is absolutely legless. In the so- 

 called round stages of Parectopa, Phy Honor ycter, Cremastobomby- 

 cia and Porphyrosela a marked tendency towards the flat type is 

 exhibited in the shape of the head-capsule. To justify his con- 

 tention, Dr. Chapman found it convenient to consider each larval 

 molt as "A separate stage of development, as distinct as is the 

 larva state from the pupa." This radical departure from all ac- 

 cepted views seems unnecessary. There is a real subdivision of 

 the larval life of the sapfeeders into distinct supernumerary stages, 

 but it is established on other grounds. The most important 

 modification induced by sapfeeding, consists in the atrophy of 

 the spinning apparatus. Whenever the habit persists through- 

 out the entire feeding period, this atrophy necessitates a special 

 stage, in which by a regeneration of the spinning organs the larva 

 is enabled to construct its cocoon. The spinning stage is as 

 highly specialized as the feeding stage. The spinneret is now the 

 organ "par excellence." In most cases maxillary and labial palps 

 are also highly developed. As the larva never feeds in this stage, 

 the mandibles become reduced. In nearly all cases, this reduc- 

 tion produces the complete inability to leave the mine for pupa- 

 tion. Various degrees of advancement in this direction may be 

 observed. Marmara has flat functioning mandibles which cross 

 each other like the blades of a pair of scissors. They are used by 

 the larva to cut its way through the epidermis of the foodplant 

 when leaving the mine and also to perforate the outer layers of 

 its cocoon to adorn it with characteristic globules in the manner 

 observed and described by Mr. Busck. Cameraria has the man- 

 dibles of a normal tissue-feeding larva, but they are much re- 

 duced and so placed as to be functionless. The mandibles of an 

 unidentified phyllocnistid (?) erroneously described as the larva 

 of Meiriochroa are merely small, shapeless chitinizations. The 

 mandibles of P.iyllocnistis seem to be completely lost. In Phyl- 

 locnistidae and Gracilariidae we thus find a series of molts so 

 specialized as to accomplish ultimately the complete separation 

 in time of two vital functions of the larva, namely that of feeding 



