232 H. M. JOHNSON AND FRANKLIN C. PASCHAL 



CONCLUSION 



The results of the tests indicate that the simpler sensorimotor 

 reflexes, as a rule, do not exhibit noticeable impairment until 

 deterioration of the more highly organized responses has reached 

 an advanced stage. Perhaps an adequate hypothetical expla- 

 nation is that the deleterious influence of low oxygen on the 

 subject's responses is largely, if not principally, due to inter- 

 ference at the synapses. On such an hypothesis one would 

 expect that the processes which first show impairment are those 

 which involve the highest degree of organization of neuromus- 

 cular arcs. 



APPENDIX 



The twenty sets of codes and test-material used in the foregoing work are re- 

 produced below. A word regarding their genesis may be useful, as a very large 

 amount of material was assembled, tried and rejected for important irregulari- 

 ties in the distribution of difficulty, before the method of compiling this material 

 was developed. The first original attempt at compilation of work-material which 

 is free from this objection will probably surprise the student by its difficulty. 



The following rules were adopted for the construction of the codes. They 

 were strictly followed, except for a few errors due to inadvertence, the correction 

 of which would have necessitated recompilation of a large part of the material. 



1. No code is systematically derived from another. 



2. Every letter in the alphabet shall be represented by a different letter in 

 the code. 



3. No letter in the code is to be substituted for its immediately preceding letter 

 in the alphabet. 



4. Habitual sequences (e.g., OK, QU, etc.), are to be avoided, as are also se- 

 quences tending to form considerable parts of words. 



5. A given letter is not to be substituted for the same letter more than twice 

 in the 20 codes. 



Subject to these limitations, the sequence of letters in each code was deter- 

 mined by a pack of shuffled cards. 



The first set of work material was prepared by shuffling a pack of 52 cards 

 containing two complete alphabets, and using the sequence of letters determined 

 in this way. 



The original set of material was then transliterated into code 1 ; and the result 

 into code 2, etc. The last two letters of each set were discarded. 



In the earliest of the experiments described in this report, some of the subjects 

 discovered that the material presented with one code, became after translitera- 

 tion the material presented with the next code. This discovery proved distract- 

 ing as the subject was tempted to compare his performance by the card, to the 

 neglect of the task. For this reason the various sets were thereafter given in 

 shuffled order, and this source of disturbance disappeared. 



