Mr. Darwin has published a short notice in reply to Professor Delpino, in _Scientific Opinion_ of October 20, 1869, p. 426. In this reply he admits the justice of Professor Delpino's attack, but objects to the alleged necessity of the first subordinate hypothesis, namely, that the emission of gemmules takes place in all states of the cell. But if this is not the case, then a great part of the utility and distinction of pangenesis is destroyed, or as Mr. Lewes justly says,[225] "If gemmules produce whole cells, we have the very power which was pronounced mysterious in larger organisms." {215} Mr. Darwin also does not see the force of the objection to the power of self-division which must be asserted of the gemmules themselves if Pangenesis be true. The objection, however, appears to many to be formidable. To admit the power of spontaneous division and multiplication in such rudimentary structures, seems a complete contradiction. The gemmules, by the hypothesis of Pangenesis, are the ultimate organized components of the body, the absolute organic atoms of which each body is composed; how then _can_ they be divisible? Any part of a gemmule would be an impossible (because a _less_ than possible) quantity. If it is divisible into still smaller organic wholes, as a germ-cell is, it must be made up as the germ-cell is, of subordinate component atoms, which are then the _true_ gemmules. This process may be repeated _ad infinitum_, unless we get to true organic atoms, the true gemmules, whatever they may be, and they necessarily will be incapable of any process of spontaneous fission. It is remarkable that Mr. Darwin brings forward in support of gemmule fission, the observation that "Thuret has seen the zoospore of an alga divide itself, and both halves germinate." Yet on the hypothesis of Pangenesis, the zoospore of an alga must contain gemmules from all the cells of the parent algae, and from all the parts of all their less remote ancestors in all their stages of existence. What wonder then that such an excessively complex body should divide and multiply; and what parity is there between such a body and a gemmule? A steam-engine and a steel-filing might equally well be compared together.