THE NATURE OE METAMORPHOSIS. 59 



chanfres in the larvai of the wild bee, AnthopJtora rctnsa, but have 

 been unable to observe them, although we believe they do really 

 occur. But the universally acknowledged accuracy of most of 

 Swammerdam's observations, supported as they are in this instance 

 by analogy, fully warrants us in considering this subject as still 

 open for inquiry. 



" When a full-grown larva is preparing to change into the pupa 

 state, it becomes exceedingly restless, ceases to eat, and diminishes 

 much in weight. Many species spin for themselves a covering of 

 silk termed a cocoon, or case, in which they await their trans- 

 formation. Others prepare little cavities in the earth, and line 

 them with silk for the same purpose ; and some suspend them- 

 selves by their hindermost legs to the under surface of a leaf. In 

 each of these instances this important change takes place in the 

 same manner. Before the larva thus prepares itself for metamor- 

 phosis, its alimentary canal is completely evacuated of its contents ; 

 its body, as at the previous changes of skin, becomes dry and 

 shrivelled, and much contracted in length; and certain enlarge- 

 ments at the sides of the anterior segments indicate the now 

 rapidly developing parts of the future pupa. These changes take 

 place in all insects in a similar manner, but have been most 

 frequently watched in the Lcpidoptcra, upon which also our own 

 observations have been made. We have also observed the same 

 changes in the Hymcnoptcra. 



" The larva of the Sphinx, when it is ready to undergo its 

 changes, penetrates the earth to the depth of a few inches, and 

 there forms for itself a little chamber, in which it awaits its trans- 

 formation. But the larva of the butterfly either fastens itself by 

 a little rope of silk carried across its thorax to the undcr-surface 

 of some object, as a ceiling, &c., or suspends itself vertically by 

 its hind legs, with its head directed downwards, as is the case with 

 the common nettle butterfly, Vanessa nrticce. We have watched 

 the changes with much care in this insect, which frequently remains 

 thus suspended for more than ten or twenty hours before the trans- 

 formation takes place. During this time the four anterior segments 

 of the larva become greatly enlarged. Then the segments assume 

 a curved direction, occasioned by the contraction or shortening of 

 those muscles of their under-surface which are repeatedly slowly 



