of Edinburgh , Session 1877 - 78 . 
585 
in the fact that r, or s, or 0, which are continuous, should sound 
alike backwards and forwards, hut it is very remarkable that sounds 
like p or k at the end of the syllable should, when produced back- 
wards, give the same effect asj? and k at the beginning of a word. 
The authors tested this singular fact in the following way : — They 
first arranged that a simple combination, such as “abafa” or 
“ afaba,” should be said to the instrument while one observer was 
out of the room. He then came in, and heard the word backwards. 
If the original word had been “ abafa,” he was in all cases able to 
say that he heard “ afaba ”; if the original word had been “ adaka,” 
he heard “ akada.” Inasmuch as he was quite uncertain whether 
the word spoken to the instrument was “ adaka ” or “ akada,” this 
experiment was in itself almost decisive. All the letters of the 
alphabet were tried in this way. Tearing, however, that when the 
alternative lay between two consonants, these might be recognised 
by some other quality than their true sound, the authors next 
proceeded to test the fact by speaking short words, such as “ amma,” 
“ abba,” and so forth, to the instrument. These were recognised by 
an observer from outside when spoken backwards. “ Alla ” was 
recognised at once; “appa” was called “abba”; “amma” was re- 
cognised at once; “ atta ” was called “adda.” 
The greatest difficulty was met with in distinguishing s from /, 
but the difficulty was not greater backwards than forwards. It 
seems due to the fact that the instrument used will not record the 
high tones which are obviously present in these sounds. 
When the fact had been established that a consonantcould be 
sounded backwards between two vowels, monosyllables were next 
tested : “ ba” was reversed and heard as “ab ;” “ ka ” was reversed, 
and heard as “ ak.” 
This is even more remarkable than the previous results, inasmuch 
as the mode of pronouncing b or k at the end of a word is different 
from that employed when it begins a word. 
Finally, by pronouncing any word backwards, such as 
“ noshaeesossa,” and then reversing the instrument, the original 
word (association) was clearly heard. With a little practice, a sen- 
tence might be spoken backwards, and then heard intelligibly 
forwards. 
The letter li was not in the phonograph to be distinguished from 
