990 



SCIENCE. 



[N. S. Vol. XVI. No. 416. 



ence without the collections; that if they are 

 made men will come to study and describe 

 them. Thousands of pages of paleontological 

 literature are worthless because paleontologists 

 have attempted to restore animals they never 

 saw from a few broken bones of the skeletons 

 —a thing that is absolutely impossible. So 

 also is it burdened with species after species 

 described from fragments, when perfect speci- 

 mens are found, many of them resolved into 

 one. Such work is worse than labor lost. 

 Nature makes no mistakes and a perfect speci- 

 men is of more value than many books describ- 

 ing poor or imperfect material. 



Charles H. Sternberg. 

 Lawrence, Kansas, 

 November 6, 1902. 



That the Carnegie Institution should, 

 above all, accomplish work not being done 

 elsewhere is the one proposition in this sym- 

 posium which has not met with disagreement. 

 The establishment of a vivarium for experi- 

 mental evolution meets this requirement. It 

 is no longer necessary to urge the value of 

 such an institution. I shall only try to 

 Bhow why work in experimental evolution 

 has been retarded and how a well-equipped 

 and well-supported vivarium will make pos- 

 sible great development in this direction. 



There is not lacking ability, interest or 

 desire to do this kind or work. It is the lack 

 of facilities which prevents effort. The re- 

 sources of the agricultural experiment sta- 

 tions are not available, because they are re- 

 stricted to economic problems. Experiments 

 in evolution are beyond the means of the un- 

 assisted worker for the following reasons : 



1. Expense. — A barn, a greenhouse and a 

 large and adequately protected garden are 

 required. Moreover, the collection of ma- 

 terial would sometimes require the expense 

 of traveling. 



2. Time. — Every-day year-arouud attention 

 is impossible for most college teachers, who 

 are generally absent from time to time for 

 lectures, meetings -and vacations, and who 

 cannot afford to employ a reliable and skilled 

 assistant to carry on the routine work in their 

 absence. 



3. Permanence. — Such experiments often 

 need to be continued through many years, 

 some even indefinitely. This is very difficult 

 in a university where such work would most 

 likely be attempted, because it would be by 

 the energy of but one man who might at any 

 time be caUed to another position. 



One must further consider that effort in 

 Buch a vivarium must be vastly more pro- 

 ductive than equal effort from individual in- 

 vestigators, for the following reasons : 



1. Division of Labor. — Since much of the 

 labor involved is of a routine nature which 

 can be carried on by a skilled gardener or at- 

 tendant, the results will be of greater value 

 because experiments involving numbers can 

 be carried out on a large scale and more prob- 

 lems can be undertaken. 



2. Superiority of Equipment. — The equip- 

 ment would be of the best, thus insuring more 

 and better results. Accidents resulting from 

 improvised or inadequate apparatus or ar- 

 rangements have spoiled many experiments ; 

 witness the work upon breeding insect larvae. 



Let us hope that the Carnegie Institution 

 will seize the opportunity of aiding in putting 

 evolution at last on an inductive basis. 



EoswELL H. Johnson. 



One respect in which German chemists have 

 an important advantage over most Americans 

 is that many of them work with the aid of 

 private assistants. These assistants are usu- 

 ally thoroughly trained men who have already 

 taken the doctor's degree. They have no 

 duties as instructors or otherwise in connec- 

 tion with teaching. 



In a few cases American chemists having 

 independent means have employed men to aid 

 them in their researches, but, with very rare 

 exceptions, colleges or universities have not 

 used their funds for such a purpose. If pri- 

 vate assistants were furnished, who should 

 work exclusively in carrying out experimental 

 researches under the direction of some of those 

 chemists who have already done good, inde- 

 pendent work, large results would, undoubt- 

 edly, be obtained. It is difficult to see how 

 in any other way so much could be accom- 

 plished toward furthering chemical research. 

 W. A. NOYES. 



