May 12, 1899.] 



SCIENCE. 



687 



admission either from Lehmann or from me. 

 Lehmanu wrote in his original paper: " Ein 

 exacter Beweis hierfiir ((. e. , for his explana- 

 tion) kaun wohl im Augenblicke nicht gefiihi't 

 werden." Nor, I take it, in any future Augen- 

 blick. 



On the other hand, I have never regarded 

 this point as the point at issue. Lehmann set out 

 to examine telepathy at large. He chose the 

 Sidgwick experiments simply as typical series, 

 considering the authors' names a guarantee of 

 serious intent and careful work. lu his inquiry 

 he laid hold of a condition which had never 

 been thoroughly investigated before, and traced 

 its effects in experiments that were both inge- 

 niously devised and rigidly controlled ; no one 

 can neglect the unconscious whisper in future 

 telepathic work. His paper is a model of 

 scientific method ; he has shown us how bor- 

 derland questions are to be attacked, and 

 proved that the ' ordinary channels of sense ' 

 have unexplored resources. His suggestions 

 will be fruitful, for the next stage of advance 

 must be an exhaustive study of the ' number 

 habits' which Sidgwick at first rejected, but 

 now makes the headstone of the corner. Even 

 granting all the contentions of the critics, there- 

 fore, I should assert that Lehmann's work is 

 brilliant, and that it has done signal service to 

 scientific psychology. But, as I hinted before, 

 I do not know that quasi-mathematics has con- 

 tributed much to psychology in any field of re- 

 search. 



I conclude with a word on the logic of Pro- 

 fessor James' objection. A theory is propounded 

 which, from the outset, lays claim to proba- 

 bility and to probability only. ' Exact proof 

 is acknowledged to be impossible. Criticism 

 plays upon the theory, and the author again 

 acknowledges that his hypothesis is not proven. 

 Professor James, apparently forgetting the first 

 acknowledgment, aflflrms that the criticism 

 has ' exploded ' the theory ! What is not proven 

 is, eo ipso, exploded ! Is Professor James, then, 

 ready to grant that his recent book on ' Human 

 Immortality ' — something which assuredly is not 

 yet proven — is an ' exploded document ' '? If 

 the alternatives before me are scientific isolation 

 and companionship on these logical terms I 

 prefer the isolation. E. B. Titchenee. 



NOTES ON PHYSICS. 



THE COMPENSATION PYEHELIOMETBR. 



Most of the measurements heretofore made 

 upon radiant energy by means of the thermopile 

 or bolometer are relative rather than absolute 

 in character, and the necessity for a simple and 

 accurate method for reducing the indications of 

 such instruments to the usual thermal units has 

 long been felt. On this account a paper by 

 Knut Angstrom {Wied. Ann., No. 3, Band 67) in 

 which he describes an instrument for measur- 

 ing radiation in absolute units is of great inter- 

 est. This instrument, to which he has given the 

 name of Compensation Pyrheliometer, is appar- 

 ently simple in construction, and the results- 

 obtained from it are very reliable, the maximum 

 error, as the author states, not exceeding 2fo. 



The construction of the instrument is briefly ' 

 as follows: Two equal, thin (.001 to .002 mm.), 

 blackened strips of platinum are mounted in 

 such a manner that either or both, by means of 

 appropriate shutters, can be exposed to the 

 radiation to be measured. 



One of the two junctions of a small constan- 

 tin-copper thermo couple is attached to each of 

 the rear surfaces of the platinum strips, the 

 circuit of thermo couple including a galvanom- 

 eter. It is evident that if one of the platinum 

 strips is exposed to radiation the equality of 

 temperature at the junctions is destroyed and 

 the galvanometer is deflected. A current of 

 electricity is now made to traverse the unex- 

 posed strip, and the strength of the current is 

 adjusted until the galvanometer returns to 

 zero. Under these conditions the two junctions 

 are receiving the same amount of energy per 

 second, and the heat developed by the current 

 in the unexposed strip is equal to that given to 

 the exposed strip by the radiation. A knowledge 

 of the strength of the current and of the resist- 

 ance of the strip sufiices to find the value of the 

 radiation in gramme calories per square centi- 

 meter per second. Since the strips are alike in 

 all respects and are subjected to identical con- 

 ditions, no corrections are necessary. 



An interesting result obtained by Angstrom 

 is the value of the mean horizontal radiation of 

 a Hefner normal lamp, which comes out to be 

 13.2 gm.-cals. per square centimeter per minute- 



