INCEPTION OF INSECT METAMORPHOSIS. 335 



pillars (Fig. >), viz., they corresponded to the normal cells of 

 tubes in the normal adult moth (Fig. C). Their height was far 

 lower than in the larval stage of development, sometimes they were 

 twice as low. Their nuclei, having a much more compact chroma- 

 tine, and a flatter, non-ramified form, were situated nearer the base 

 of the cells. The interior margin of the tubes was not so much 

 notched as in the caterpillar. All these characters prove that the 

 cells of the implanted tubes have undergone normal metamor- 

 phosis. This metamorphosis of the Malpighian tubes is conse- 

 quently independent of the function of this organ, as it is difficult 

 to suppose that the tubes undergoing metamorphosis in hetero- 

 geneous surroundings were able to realize their special function in 

 the chrysalis and later in the adult moth, as they were unconnected 

 with the developing intestine, completely grown together and al- 

 ways filled with excrements previously formed during the larval 

 life. If it could be proved without any objection that the Mal- 

 pighian tubes which have been implanted into the head or thorax 

 had no nervous junction in the new surroundings, the metamor- 

 phosis of transplanted fragments of Malpighian tubes would be a 

 further proof for the assertion made in the previous chapter that 

 the brain exerts its influence by means of internal secretion. The 

 probably unchanged rate of metamorphosis of the tubes transferred 

 from caterpillars 10 days after their last moult into those 2 to 3 

 days after this moult (thus from specimens in which the influence 

 of the brain was already observable into specimens in which there 

 was yet no such influence) points to the conclusion that tissues once 

 stimulated by the brain to transformation undergo further meta- 

 morphosis independently. 



Should this principle be of general importance, it would be ex- 

 pected that the organs of much older caterpillars (a few days 

 before pupation) grafted on the much younger caterpillars after 

 the fourth moult (and therefore some two to three weeks younger) 

 might also undergo further metamorphosis at the normal rate, in 

 spite, of the fact that the organs of the younger specimens do 

 not as yet indicate the slightest trace of the histolytical processes. 

 Transplantations of the germ of wings fully confirmed this hy- 

 pothesis. I removed the germ of the first left wing from numer- 

 ous caterpillars shortly before their last moult, and in its place I 



