From the evidence that his results have given, De Vries believes it is probable that mutation has occurred in all directions. In the same way that Darwin supposed that individual or fluctuating variations are scattering, so also De Vries believes that the new forms that arise through mutation are scattering. On this point it seems to me that De Vries may be too much prejudiced by his results with the evening primrose. If, as he supposes, many forms, generally ranked as varieties, are really elementary species, it seems more probable that the mutation of a form may often be limited to the production of one or of only a very few new forms. The single variations, or sports, point even more strongly in favor of this interpretation. Moreover, the general problem of evolution from a purely theoretical point of view is very much simplified, if we assume that the kinds of mutating forms may often be very limited, and that mutations may often continue to occur in a direct line. On this last point, De Vries argues that the evidence from paleontology cannot be trusted, for all that we can conclude from fossil remains is that certain mutations have dominated, and have been sufficiently abundant to leave a record. In other words, the conditions may have been such that only certain forms could find a foothold. De Vries asks whether there are for each species periods of mutation when many and great changes take place, and periods when relatively little change occurs. The evidence upon which to form an opinion is scanty, but De Vries is inclined to think that such periods do occur. It is at least certain from our experience that there are long periods when we do not see new forms arising, while at other times, although we know very few of them, epidemics of change may take place. The mutative period which De Vries found in the evening primrose is the best-known example of such a period of active mutation. Equally important for the descent theory is the idea that the same mutation may appear time after time. There is good evidence to show that this really occurs, and in consequence the chances for the perpetuation of such a form are greatly increased. Delbœuf, who advocated this idea of the repeated reappearance of a new form, has also attempted to show that if this occurs the new form may become established without selection of any kind taking place,—the time required depending upon the frequency with which the new form appears. This law of Delbœuf, De Vries believes, is correct from the point of view of the mutation theory. It explains, in a very simple way, the existence of numerous species-characters that are entirely useless, such, for instance, as exist between the different elementary species of _Draba verna_. “According to the selection theory only useful characters can survive; according to the mutation theory, useless characters also may survive, and even those that may be hurtful to a small degree.”