CONSTRUCTION OF THE CATERPILLAR. 137 



rudiments of a leaf or of a flower in the bud, that 

 Swammerdam has given figures of the parallel deve- 

 lopments of larvae and of a carnation. His selection 

 of this flower was not perhaps the most happy ; but 

 our readers may readily obtain examples by carefully 

 dividing the unexpanded buds of the rose, the lilac, 

 the horse-chesnut, the American walnut, or beans, 

 and other large seeds after they have been planted in 

 moist earth, but not left long enough to shoot into a 

 plant. The preceding figures will illustrate this better 

 than description. Dr. Grew proved in this manner 

 that flowers which blow in spring are formed in the 

 preceding year*; and Du Hamel, on dissecting, in 

 January, the bud of a pear-tree, found under an 

 envelope of about thirty leaf-scales eight or ten 

 embryo flowers resembling rose-buds bestudded with 

 hairs f. 



The butterfly and the flower-bud, however, differ 

 remarkably in the manner in which they are nourished, 

 — the latter receiving sap from the enveloping leaf- 

 scales, the former taking food into the stomach through 

 the mouth of the caterpillar. The stomach, indeed, 

 of the inclosed butterfly is so capacious, that it fills 

 the greater portion of its body ; and requires the cater- 

 pillar to occupy almost its whole time in eating in 

 order to satisfy its cravings. When the food is digested 

 in the stomach of the insect, it passes, as in the larger 

 animals, into the small intestines J : but it is not, as in 

 them, collected by innumerable little vessels which 

 afterwards run into one, (as brooks unite to form a 

 river,) and go to the lungs to be exposed to the air 

 supplied by breathing, in order to be there oxygenated 

 and formed into red blood ; insects, on the contrary, 



* Gicw. Phys. Veg., ii. 60. 



t Du Hamel, Physique ties Arbres, iii. 1. 



t See * Insect Architecture,' p. 309, D.D.; and this vol. p. 198, 



i3 



