February 22, 1906] 



NA TURE 



405 



it was possible to establish normal chlorine, or, as they 

 are commonly called, iso-chlor lines. 



The work begun by Prof. Drown has been carried on 

 by other investigators, and to-day the iso-chlor lines for 

 all the New England States and New York and New Jersey- 

 have been determined. The result of this work is that 

 the amount of chlorine occurring in the surface waters ol 

 the above named States gives most valuable information. 

 Chlorine above the normal of the region shows pollution. 

 It does not indicate whether the pollution is direct or 

 indirect, but does show that sewage, from which the 

 organic matter and the germs of disease may or may not 

 have been removed by filtration through soil, has had 

 access to the water. Chlorine above the normal is, there- 

 fore, always a suspicious sign which must be investigated. 

 1 know that it is claimed that in many of the western 

 States, owing to geological conditions, very little inform- 

 ation can be obtained from the determination of chlorine. 



I believe, however, more careful and thorough work is 

 necessary to prove that such is the case, and that further 

 investigation may show that though it is impossible to 

 construct iso chlor lines running through the State, the 

 normal chlorine of different localities in a State can often 

 be determined. 



Another factor that is often used in the attempt to decide 

 whether or not a water contains an excessive amount of 

 organic matter is the oxygen consumed. The oxygen con- 

 sumed is not, however, a measure of the organic matter 

 in a water, but only a measure of the amount of mineral 

 reducing salts plus a certain amount of the organic matter, 

 the amount depending on the method of determination used. 



I I gives, in my opinion, very little information as to the 

 character of the organic matter, and is only valuable when 

 different surface waters are to be compared with each 

 other, or when used in filtration experiments. 



The same may be said as regards colour, turbidity, and 

 the amount of mineral matter that a surface water con- 

 tains, that, though of essential importance in deciding on 

 the value of a normal water as a potable water, they give 

 little information as to pollution. 



In the early days of bacteriology it was claimed that the 

 final criterion as to pollution of a water would be furnished 

 by aid of that science, and though this hope has not been 

 fulfilled, the information that can be gained by a bacterial 

 analysis is often of the highest importance. It not only 

 aids in the interpretation of the chemical data, but may 

 of itself show, almost without question, that a given water 

 is polluted, for though attempts to isolate special patho- 

 genic germs have generallv failed, even in waters known 

 to contain these forms, characteristic sewage forms, like 

 the colon bacillus, can be isolated if they occur in any 

 number in a water. Occurrence of numerous characteristic 

 sewage bacteria can point only to one thing, pollution, and 

 if such forms are found there is no question that the water 

 receives sewage drainage. Bacteriology, however, cannot 

 determine, except very roughly, the amount of pollution 

 or the present condition of the polluting matter, nor does 

 it give but very little, if any, information as to past pollu- 

 tion. If the pollution is recent and of any considerable 

 amount, a careful bacterial examination will show the 

 fact, and probably better and more convincingly than any 

 chemical analysis. If the pollution is more remote, more 

 information can, as a rule, be drawn from chemical than 

 from bacterial data. If the polluting matter has filtered 

 through the soil before entering the water, bacterial work 

 will not indicate the fact. 



As a general statement, it may be said that a bacterial 

 analysis, while- giving information as regards recent and 

 continuous pollution, gives no information as to the past 

 history of a water, and in this respect differs from a 

 sanitary chemical analysis. 



To form a judgment as to the wholesomeness of a 

 water, the data of a sanitary water analysis, the source 

 of the water, whether surface, ground, r.r artesian, must 

 be known ; a survey, even of a surface water, though 

 it may show whether or not the water is polluted, does 

 ■11 give information regarding the amount or condition of 

 the polluting matter ; with ground and artesian waters 

 it often gives very little information, and an opinion 

 regarding the character of such waters must, as a rule, 

 depend on the sanitary analysis. 



NO. 1895, VOL. 73I 



The Generic Concept in the Classification of the 

 Flowering Plants. 1 



The difficulties of classifying plants in a really natural 

 and logical way are somewhat increased by the involuntary 

 and well-nigh necessary admission of a certain historic 

 element into our systems. There is another source of 

 this artificiality, besides the temptation to allow poor genera 

 to stand, on the ground of long usage. The relation of a 

 genus to its name is a matter which exerts no small in- 

 fluence in this regard. The attempt to determine which 

 of several names is to be retained for a given genus con- 

 stantly forces us to consider the historic basis on which 

 the genus rests, and to attach its name to some species 

 or group of species to which it was first applied, to deter- 

 mine, in other words, what was the type of the genus, and 

 to maintain the genus in such a way that it may always 

 be true to its type. While sympathising to a considerable 

 extent with those botanists who desire to place our nomen- 

 clature upon a more secure basis by attaching the names 

 to recognised types, I feel that the methods employed will 

 have to be very cautiously applied, or they will tend 

 greatly to increase the artificial element in our system. 

 The historic type is not a natural thing ; it is merely that 

 particular form of plant life which was, often quite by 

 accident, first discovered and, therefore, first received the 

 name which it bears. Later discoveries often show that 

 this first species of a genus is by no means of a typical, 

 or, as one may say, central character. It is often quite 

 peripheral, perhaps even an aberrant or outlying member 

 of the group to which it belongs. However important the 

 historic type may be in nomenclature, it is obvious that 

 it is of no particular significance in classification, and any 

 employment of the type method in the determination of 

 proper names must not on any account be permitted to 

 exercise any influence in classification. The word type 

 itself is decidedly unfortunate as thus applied to what is 

 often very far from being typical. In this as in some 

 other phases of taxonomy it is of the greatest importance 

 to keep it clearly in mind that nomenclature, although very 

 necessary to classification, is a thing wholly apart from 

 the classification itself. It is, furthermore, quite evident 

 that nomenclature should be subservient to classification, 

 and that the clearness and accuracy of classification should 

 never be sacrificed in order to give beauty or symmetry 

 to anv system of nomenclature. 



The limitation of genera has always in the past rested 

 on individual judgment, and it must continue to do so in 

 the future. Although the genera of the flowering plants 

 have now been scientifically studied for about two cen- 

 turies, there is at present in America, at least, a degree 

 of diversity in their interpretation which is rather dis- 

 couraging. It is disheartening because it is impossible to 

 see in it any real progress toward a well rounded and 

 satisfying system, which will win the confidence of the 

 professional botanist, give uniform training to the student, 

 and command the respect of our colleagues in other 

 branches of science. From this, I think that it is per- 

 fectly clear that botanical systematists have certain im- 

 perative duties in regard to this subject. These duties are, 

 in the first place, great caution in making changes, and, 

 in the second place, a feeling of obligation, when these 

 changes seem necessary, to state the reasons for them so 

 clearly and forcibly that they will appeal to all thoughtful 

 and discriminating workers in the same field. The burden 

 of proof should always rest upon the writer suggesting 

 the change. 



What we need in botanical classification is a series of 

 constants, a number of graded categories which can be 

 generally endorsed and properly respected. Standards as 

 definite as those of the physicist are, of course, quite 

 unattainable in dealing with the variable and often inter- 

 grading groups of organic creation. But where absolute 

 accuracy and uniformity are impossible, we should the 

 more diligently seek to preserve such standards as exist. 

 As has been pointed out, there are but few lamilies of 

 flowering plants which have not been comprehensively 

 treated by monographers who, so far as their particular 

 group was concerned, have been in a position to see pretty 



1 From the address o. Prof. B. L. Robinson, president of Section G, 



