Differentiations thus arising, both before and after separation from parents, though primarily differences of growth, entail structural differences; for it is a general law of nutrition that when there is deficiency of food the non-essential organs suffer more than the essential ones, and the unlikenesses of proportion hence arising constitute unlikenesses of structure. It may be concluded, however, that variations generated in this manner usually have no permanent results. In the first place, the individuals which, primarily in growth and secondarily in smaller developments of less-important organs, are by implication inferior, are likely to be eliminated from the species. In the second place, differences of structure produced in the way shown do not express differences of constitution--are not the effects of somewhat divergent physiological units; and consequently are not likely to be repeated in posterity. § 88. We have still, therefore, to explain those variations which have no manifest causes of the kinds thus far considered. These are the variations termed "spontaneous." Not that those who apply to them this word, or some equivalent, mean to imply that they are uncaused. Mr. Darwin expressly guards himself against such an interpretation. He says:--"I have hitherto sometimes spoken as if the variations--so common and multiform in organic beings under domestication, and in a lesser degree in those in a state of nature--had been due to chance. This, of course, is a wholly incorrect expression, but it serves to acknowledge plainly our ignorance of the cause of each particular variation." Not only, however, do I hold, in common with Mr. Darwin, that there must be some cause for these apparently-spontaneous variations, but it seems to me that a definite cause is assignable. I think it may be shown that unlikenesses must necessarily arise even between the new individuals simultaneously produced by the same parents. Instead of the occurrence of such variations being inexplicable, the absence of them would be inexplicable.