Apbil 29, 1910] 



SCIENCE 



677 



stout dehiscent pods. Acacia cornigera L. differs 

 from both of these in having inflated indehiscent 

 pods terminating in a spine-like beak, as well as 

 in the character of its inflorescence and of the 

 extrafloral nectaries on its leaves. 



Acacia cookii sp. nov. Flowers in spherical 

 heads on long stout peduncles clustered in the 

 axils of large slender thorns resembling the prongs 

 of a fork which usually straddle the stem ; leaves 

 large, with many pairs of pinnse and many elon- 

 gated nectar glands borne on the upper side of 

 the grooved rachis; pods linear, 30 cm. or more 

 in length, slightly curved, dehiscent. Based on 

 specimens collected by Mr. O. F. Cook at Secan- 

 quim, Alta Verapaz, Guatemala (in alcohol), and 

 by Mr. G. P. Goll, at the Finea Trece Aguas, in 

 the same region, March 8, 1907 (No. 102). 



Acacia collinsii sp. nov. Flowers in spadix-like 

 spikes, usually in clusters of four or five, the 

 oldest spike usually sessile or nearly so, the rest 

 on long stout peduncles; bractlets of the inflores- 

 cence peltate circular, covering the unopened 

 flowers, but concealed after anthesis; leaves with 

 several round bead-like nectar-glands at the base 

 of the petiole and a single gland on the rachis at 

 the base of each pair of pinnae; thorns stout, 

 U-shaped; one of the arms usually perforated by 

 ants, as in the case of other " bull-horn " acacias ; 

 pods stout, thick, short, straight or slightly 

 curved, dehiscent, filled with yellow sweetish aril 

 in which the seeds are imbedded. This species is 

 based oa specimens collected by Mr. Guy N. Col- 

 lins between Chicoasen and San Femandino, in 

 the state of Chiapas, southern Mexico, January 

 14, 1907 (No. 180). A species resembling Acacia 

 hindsii, but differing from that species in the 

 form of its thorns, the thickness of its peduncles, 

 and the form and stoutness of its pods. 



The Categories of Variation: W. J. Spillman, 



U. S. Bureau of Plant Industry. 



Recent work indicates that the variations with 

 which Darwin dealt may be separated into sev- 

 eral categories which have different relations to 

 evolutionary change. The work of Nilsson, 

 Johannsen and Jennings seems to have demon- 

 strated that there is a class of variations, due 

 wholly to environment, that are not hereditary 

 and on which selection is without effect. These 

 variations are coming more and more to be called 

 " fluctuations." 



It is also pretty well established that when an 

 organism is removed from its old environment to 

 an entirely new one it may undergo rather marked 

 changes, apparently as the result of changed en- 



vironment. Tlie meager information at hand indi- 

 cates that several individuals having exactly the 

 same inheritance undergo the same change when 

 transplanted to a new environment and that the 

 change is permanent under the new environment. 

 Some recent investigations indicate that in cases 

 of this kind, when the organism is transferred 

 back to its old environment, it changes back to 

 its old form. Much more investigation is needed 

 before this type of variation, which is sometimes 

 called " new-place effect," can be properly cata- 

 logued. 



A third type of variation is that due to re- 

 combination of Mendelian characters. These re- 

 combinations frequently result in the production 

 of new forms which are stable and must therefore 

 be looked upon as one means of progressive 

 evolution. 



Apparently a fourth type of variation is that 

 discovered by de Vries in (Enothera. The investi- 

 gations of Gates and Miss Lutz point to the as- 

 sumption that the variations studied by de Vries 

 are due to the loss, gain or exchange of chromo- 

 somes in mitosis. 



There are probably many other types of varia- 

 tion which have not yet been recognized. On 

 a priori grounds it would appear almost certain 

 that changes in the chemical composition of the 

 germ plasm or in the relative amounts of sub- 

 stances present in the germ plasm are of funda- 

 mental importance in evolution, and that in the 

 main evolutionary progress is due to them. These 

 changes may take place in any part of the germ 

 cell which has a determining influence on devel- 

 opment. It was suggested that when such a 

 change occurs in the composition of a chromosome 

 the new form resulting would give Mendelian 

 phenomena when crossed with the old form, but 

 if the change occurs in cytoplasm Mendelian phe- 

 nomena would be lacking, and there is some evi- 

 dence that this is the case. A case in point is that 

 of albomarginate leaves studied by Baur. The 

 behavior of the cross is such as to indicate that 

 the albomarginate character is cytoplasmic, and 

 the inheritance of this character is non-Mendelian. 

 W. W. Stockbebgee, 

 Corresponding Secretary 



THE ANTHEOPOLOGICAl SOCIETT OP WASHINGTON 



At the 445th regular meeting, held March 29, 

 1910, the first paper of the evening was on " The 

 White-dog Feast of the Iroquois," by Mr. J. N. 

 B. Hewitt. 



The white-dog sacrifice of the Iroquois is a 



