May 23, 1913] 



SCIENCE 



795 



mineral matter. Moreover, altliougli it is rela- 

 tively easy to distinguisli between CO, and 

 H„0 on the one hand as inorganic " food 

 materials," and sugars, starches, etc., on the 

 other hand as manufactured " foods," who can 

 say when nitrogen, sulphur and phosphorus 

 cease to be " food materials " and become 

 " foods " ? Is it not more than probable, also, 

 that some constituents of the mineral material 

 taken in by plants and animals are immedi- 

 ately available for assimilation in the form 

 absorbed, and are thus foods in both the re- 

 stricted and broader senses of the word? If 

 the facts are as here suggested, it is clearly 

 impossible to limit the term food to organic 

 material, first because too little is known of 

 the metabolic processes by which nitrogen, 

 sulphur, phosphorus, et al., are assimilated to 

 enable any one to say at what stage these 

 elements cease to be parts of inorganic and 

 become parts of organic compounds, and sec- 

 ond, because some inorganic substances are 

 probably foods in both senses of the word. 



In conclusion, the question asked in the 

 title may be repeated. How is the word food 

 to be defined? Is it to be limited to organic 

 substances with all the pedagogic and scientific 

 difficulties which such limitation entails? Or 

 shall it remain as at present, raising no prac- 

 tical difficulties whatever and leaving the 

 academic difficulties involved to be dealt with 

 when the pupil becomes sufficiently mature to 

 imderstand them? Ealph C. Benedict 



High School op Commerce, 

 New York City 



a standard foem of committee meetings 

 Distances in the United States are so great 

 that it is often impossible for a committee to 

 hold a meeting, and its work must be done by 

 correspondence. Owing to the international, 

 or national, character of many committees, 

 and the increasing amount of friendly coop- 

 eration among scientific men, some standard 

 system of arriving at results is greatly needed. 

 It would then only be necessary when appoint- 

 ing a committee to state that its work would 

 be done in this way, and the chairman would 

 be saved the necessity of devising a method in 



each case, and the doubt in many cases, 

 whether he was justified in appending the 

 names of all members to his report. 



The following method is accordingly sug- 

 gested: The chairman or secretary should 

 have three letters manifolded, and sent in due 

 course to each member of the committee. The 

 first of these should state the exact terms of 

 the appointment; the objects desired; a re- 

 quest for suggestions for the report; an opin- 

 ion whether a meeting is advisable, and if so, 

 when and where. 



The second letter should contain a prelim- 

 inary report embodying the suggestions re- 

 ceived, and in cases of doubt asking numbered 

 questions to which, if possible, the answer will 

 be yes, doubtful, disapprove, or no. In the 

 first three cases, the writer accepts the views 

 of the majority of the committee. In all four 

 cases, he authorizes his name to be attached to 

 the report, provided that it contains a state- 

 ment that he dissents from the questions to 

 which his reply is no. Prompt answers are 

 requested, but if any member fails to reply 

 after a letter has been in his hands for a week, 

 the chairman may assume that he assents. 



The third letter should contain the proposed 

 report, to which all the names would be at- 

 tached unless answers were received express- 

 ing dissent. Some of the members might 

 prefer to make a minority report. If no 

 reply was received to letters one and two, 

 letter three should be registered with a request 

 for a receipt, as otherwise the previous letters 

 might not have been received. If haste is 

 important, night letters are generally to be 

 preferred to telegrams, since the delay from 

 the most distant points of the country would 

 seldom exceed twenty-four hours. If a reply 

 by cable is necessary, the chairman should 

 give his cable address, and if possible arrange 

 all his questions so that each answer shall 

 consist of only one or two words. A reply by 

 cable, in which the fifth question related to a 

 place of meeting, might read: Fieldsmith, 

 Washington. One, two, yes; three, no; four, 

 doubtful; five, London, July. Brown. 



Edward C. Pickering 



May 6, 1913 



