We must distinguish between hybrids proper and crosses, as it were between races or between what may have been erroneously designated species, for there has been a great deal of looseness in the manner of using these terms by some writers. A true _hybrid_[9] is produced only when the pollen of one species has been used to fertilize the ovules of another, and as a general rule these can only be produced between plants which are very nearly allied, as between species of the same genus. Even such as these, however, cannot always be hybridized, for we have never found a mule or hybrid between the apple and pear, the currant and gooseberry, nor between the raspberry and blackberry, though each of these, respectively, appear to be very nearly related, and they are all of the order _Rosaceæ_. In hybrids there appears to be a mixture of the elements of each, and the characters of the mule or cross will depend upon one or the other, which it will more nearly resemble. True hybrids are mules or infertile, and cannot be continued by seed, but must be propagated by cuttings, or layers, or grafting. If not absolutely sterile at first, they become so in the course of the second or third generation. This is proved by several of our flowering plants that have been wonderfully varied by ingenious crossing of different species. But it has been found that the hybrid may be fertilized by pollen taken from one of its parents, and that then the offspring assumes the characters of that parent.[10] Natural hybrids do not often occur, though in dioecious plants, this seems to have been the case with willows that present such an intricate puzzle to botanists in their classification, so that it has become almost impossible to say what are the limits and bounds of some of the species. Hybrids are, however, very frequently produced by art, and particularly among our flowering plants, under the hands of ingenious gardeners. Herbert thinks, from his observations, "that the flowers and organs of reproduction partake of the characters of the female parent, while the foliage and habit, or the organs of vegetation, resemble the male."