Moreover, not only from all these, but from each and every successive stage of the evolution and development of such successively more and more elementary parts.

At the first glance this new atomic theory has charms from its apparent simplicity, but the attempt thus to follow it out into its ultimate limits and extreme consequences seems to indicate that it is at once insufficient and cumbrous.
Mr.Darwin himself is, of course, fully aware that there must be _some_ limit to this aggregation of gemmules.

He says:[220] "Excessively minute and numerous as they are believed to be, an infinite number derived, during a long course of modification and descent, from each cell of each progenitor, could not be supported and nourished by the organism." But apart from these matters, which will be more fully considered further on, the hypothesis not only does not appear to account for certain phenomena which, in order to be a valid theory, it ought to account for; but it seems absolutely to conflict with patent and notorious facts.
How, for example, does it explain the peculiar reproduction which is {211} found to take place in certain marine worms--certain annelids?
[Illustration: AN ANNELID DIVIDING SPONTANEOUSLY.
(A new head having been formed towards the hinder end of the body of the parent.)] In such creatures we see that, from time to time, one of the segments of the body gradually becomes modified till it assumes the condition of a head, and this remarkable phenomenon is repeated again and again, the body of the worm thus multiplying serially into new individuals which successively detach themselves from the older portion.

The development of such a mode of reproduction by "Natural Selection" seems not less inexplicable than does its continued performance through the aid of {212} "pangenesis." For how can gemmules attach themselves to others to which they do not normally or generally succeed?
Scarcely less difficult to understand is the process of the stomach-carrying-off mode of metamorphosis before spoken of as existing in the Echinoderms.

Next, as to certain patent and notorious facts: On the hypothesis of pangenesis, no creature can develop an organ unless it possesses the component gemmules which serve for its formation.