436 



SCIENCE. 



[N.S. Vol. XVIII. No. 457. 



denied all pretense to finality of conclusions; 

 it was to get light on just those subtile changes 

 of soil, etc., to which Mr. Daniels alludes, 

 that the work was undertaken. Thus the 

 criticism that we can not point out any single 

 cause for any difierence in vegetation is 

 hardly to the point. My critic deals only 

 with the societies, he is willing to leave the 

 description of soils as it was published. 



That the societies used are not of equal 

 rank was pointed out by the author. In- 

 creased complexity would doubtless be ac- 

 companied in some measure by increased ac- 

 curacy, but the main points would thus be 

 almost surely lost sight of in the maze of 

 classification. The systematist bom and bred 

 can seldom understand the horribly slipshod 

 ways of the ecologist! The fact that Mr. 

 Daniels puts the beech-maple and the maple- 

 elm-agrimony societies together shows clearly 

 that he has failed to study the southern town- 

 ships. Farther north these types do tend to 

 merge, but southward, into Indiana and Illi- 

 nois, they are still distinct, as more recent 

 studies by -the author show. Likewise, the 

 combining of the oak-hickory with the oak- 

 hazel society indicates failure to study well 

 the condition of things in the northern part 

 of the county. The presence or absence of 

 hickory in this region involves one of the 

 most important ecological questions, one which 

 we can not interpret clearly as yet, but one 

 which it can do no good to slur over. On the 

 whole, my critic's description agrees fairly 

 well with the conditions in the central town- 

 ships, but fails altogether to express the facts 

 throughout the county as a whole. This is 

 one of the dangers which he might be warned 

 against, namely, that of generalizing from too 

 small an area, no matter how many years he 

 has botanized in it. The author must admit 

 that he, too, is a native of Kent County, and 

 that he has studied its flora for many years. 



Bidens frondosa and Solanum nigrum are 

 ' frequent everywhere ' only in the moister 

 soils of the Grand Eiver valley. Nepeta, Phy- 

 tolacca and Euphorbia corrallata are, through- 

 out the area, among the most constant and 

 characteristic members of their societies; 



they are not weeds excepting in areas where 

 they originally occurred. Echinospennum oc- 

 curs in the lowlands, a fact which may have 

 misled Mr. Daniels, but on the uplands it is 

 a truly characteristic plant of the most meso- 

 phytic society. Regarding the distribution of 

 Cenchrus there is need of more study than 

 either Mr. Daniels or myself has yet accom- 

 plished. Wherever it comes from, it certainly 

 occurs only on the worst sands, i. e., only in 

 the most xerophytic group. Occasional plants 

 of Dracocephalum are to be found in the 

 beech and maple forests south of the Grand 

 River valley. 



The societies are based on forms which are 

 evident the year round, not on summer forms 

 nor spring forms. Of the 140 forms listed, 

 no less than 40 are spring-flowering. There 

 doubtless are well-marked societies of Thallo- 

 phytes and Bryophytes, but those of the two 

 higher groups are the ones chosen for study. 

 An adequate study of the distribution of 

 grasses and sedges will be a work by itself 

 when the right student undertakes it. These 

 plants were purposely omitted from the paper 

 under discussion. 



Burton Edward Livingston; 



The New Yobk Botanicai, Garden, 

 September 15, 1903. 



SHORTER ARTICLES. 



FOUR NEW SPECIES OF THE CENTRAL AMERICAN 



RUBBER TREE. 



Though still in the initial and experimental 

 stage, the cultivation of rubber-producing 

 trees is now attracting more general atten- 

 tion than any other branch of tropical agri- 

 culture. Large amounts of American capital 

 are being invested in Mexico and Central 

 America, and the practicability of rubber cul- 

 ture in the tropical islands of the United 

 States is receiving the attention of the De- 

 partment of Agriculture. The first studies 

 have been directed to the Central American 

 rubber tree (Castilla), and one of the facts 

 established is the existence of several differ- 

 ent local types, instead of a single species ex- 

 tending from Mexico to Bolivia, as hitherto 

 supposed. The species of Castilla are among 



