August 6, 1915] 



SCIENCE 



193 



marked. One speaker in particular, as the re- 

 sult of several hearings, in the first half-year 

 was rated at 10 per cent., in the second at 76 

 per cent. 



Of the different speakers, six were heard in 

 both of the semesters, and are thus more di- 

 rectly comparable than the others. Their aver- 

 age for the first semester was 64 per cent., for 

 the second, 92 per cent., that is, while during 

 the first semester more than one word out of 

 every three was unintelligible, only about one 

 in twelve was unheard in the second. 



The effect was even more striking in seat 

 V. Before the treatment of the chapel the 

 average audibility was 71 per cent., exactly 

 the same as at the greater distance, showing 

 that the advantage gained by a somewhat 

 nearer approach to the speaker was completely 

 nullified by the disturbances from reverbera- 

 tion. The attention, as in the seat AA, was 

 careful or strained. After the treatment the 

 average audibility rose to nearly 96 per cent., 

 nearly perfect hearing, and the attention in 

 most cases was noted as easy. 



In seat Q, about fifty-five feet from the 

 speaker, the audibility rose from an average of 

 95 per cent, in the first semester to 100 per 

 cent, in each separate case in the second. 



The results are summarized below for more 

 easy comparison. 



AVERAGE OF ALL SPEAKERS 



First Sem- Second Sem- 

 ester. Per ester, Per 

 Seat Ceut. Cent. 



AA 71 91 



V 71 96 



Q 95 100 



The seat in the gallery gave exactly similar 

 results, but the number of experiments made 

 in this seat was so small that the averages are 

 not included in the table. 



The condition of the auditorium at present 

 is satisfactory. It is quite possible that a 

 slight further reduction of reverberation might 

 be made with advantage to the spoken word, 

 but the effect of music, which forms an impor- 

 tant part of the uses of the building, would be 

 correspondingly injured. 



It may be worth while further to remark 



that the calculations as to the effect of rever- 

 beration could have been as well made, plan 

 and materials being given, before the erection 

 of the building as afterward. It is a pity that 

 architects still construct buildings of this kind 

 without giving careful attention to their ex- 

 pected uses, trusting to good fortune for acous- 

 tic fitness which might easily and certainly be 

 insured in advance. 



FRAjfK P. Whitman 

 Physical Laboratory, 

 Western Reserve University 



SPECIAL ABTICLES 



two color mutations op rats which show 

 partial coupling 



In the American Naturalist for February, 

 1914, Castle described two yellow-coated vari- 

 eties of the Norway rat (Mus norvegicus) 

 which had recently been discovered in Eng- 

 land, and both of which had been found to 

 behave as Mendelian recessive characters in 

 heredity. One of these was called " pink-eyed 

 yellow," the other " black-eyed yellow." A 

 more appropriate name for the latter would be 

 " red-eyed yellow " (which we shall hereafter 

 use), since the eyes in this variety are not as 

 dark as in wild gray or tame black rats, but 

 the red blood of the eye shows through, par- 

 ticularly when the animal is young, giving 

 the eye in a favorable light a reddish tinge. 



Upon crossing the two yellow varieties with 

 each other, we found them to be complemen- 

 tary. The Fj young obtained were none of 

 them yellow, but were all either gray or black 

 coated; yet it should be noted that they were 

 in no case as dark as ordinary gray or black 

 rats. Nevertheless F^ young with coats of nor- 

 mal intensity were later obtained, so that the 

 paleness of the Fj young was evidently due 

 rather to their being heterozygous for the two 

 complementary factors, than to any failure of 

 one variation completely to supply what was 

 lacking in the other. 



Each of the yellow varieties was also found 

 to be different in nature from ordinary albin- 

 ism, as seen in white rats, since when it was 

 crossed with albinos it produced only fully 



