————-— __——————' __———- __———' ._———— __;———- ’- ————- —_’-— ’— _————- _—'— ’— _—————- —_——-——— ANXIETY AFTER AVOIDANCE LEARNING AS MEASURED BY GASTRIC SECRET iONS Thesis ‘or ”10 Degree of M. A. MICHIGAN STATE UNIVERSITY Ted Tetzlaff 1961 ANXIETY AFTER AVOIDANCE LEARNING AS MEASURED BY GASTRIC SECRETIONS BY TED TETZLAFF A THESIS Submitted to the College or Arte and Science: of Michigan State University of Agriculture and Applied Science in partial fulfillment of of the requirements for the degree of MASTER.OF ARTS Department of Psychology 1961 ABSTRACT The purpose of this study is to test, through an ob- Jective physiological measure, Denny's "incubation of anxiety" hypothesis advanced to account for the unusual re- tention of an incompletely learned avoidance response. Free HGl secretion in the stomach was used as a measure of anxiety. Three experimental groups of rats were given 25 avoidance learning trials in a shuttlebox followed by a second set of 25 relearning trials at intervals of 1 hour, 6 hours, and 24 hours, respectively. One hour prior to the second set of trials, the surgi- cal Operation of pylorus ligation was performed as a means of collecting the gastric secretions, including free H01. A control group which did not receive any avoidance trials was pylorus ligated and used to establish a basal level of H01 secretions. The results of the behavioral data were consistent with previous findings which had been obtained under similar condi- tions. The results of the analysis of free HCl yielded no significant differences among groups. However, the trend of the means for the amount of free H01 was in approximate alignment with the predictions from Denny's "incubation of anxiety" hypothesis. Approved: mg ' m Date: kW 1% (75/ ‘7 Majofi Professor To Mar: ie ACKNOWLEDGMENT The author wishes to express his sincere thanks and appreciation to the committee chairman, Dr. S. Ratner, whose guidance and encouragement aided so much in the completion of this study. A.special debt of gratitude is also owed to the late Dr. J. P. King for his assistance and advice in the planning and initial execution of this research. The author wishes to convey his thanks to Dr. Terrence M. Allen, Dr. Charles Henley, and Dr. M. Ray Denny for their helpful suggestions of the analysis of the data. The cooperation of Dr. R. K. Ringer of the Department of Poultry Science in providing necessary equipment and the invaluable assistance of the author's wife, Marjorie, in con- ducting the gastric analysis is also gratefully appreciated. 111 LIST OF TABLES INTRODUCTION METHOD . . RESULTS . . DISCUSSION SUMMARY . . APPENDICES REFERENCES TABLE OF CONTENTS iv Page 10 13 17 24 27 Table 2. 3. LIST OF TABLES Experimental procedure and treat- ment for the experimental groups and control group . . . . . . . . Means and standard deviations of the number of avoidances on original and relearning trials for the experimental groups . . . Analysis of covariance of the avoidance scores for the experimental groups . . . . . . . Observed and adjusted mean scores for the number of avoidances on relearning trials . . . . . . . . Means of the physiological mea- sures for the experimental groups and the contml group e e e e e e Standard deviations of the physio- logical measures for the experi- mental groups and the control groupeeooeeeeeeeeoe Summarized data of avoidance learning and relearning for the experimental groups . . . . . . . Summarized data of total volume and free HCl for the experi- mental groups and control group . Page 19 2O 21 22 23 23 25 26 INTRODUCTION Investigating the retention of an incompletely learned avoidance response, Kamin (1957) obtained results which pose some interesting problems of interpretation. In Kamin's pro- cedure hooded rats were given 25 ayoidance learning trials in a shuttlebox. Retention of the avoidance response was mea- sured in 25 additional learning trials after delays of 0, .5, 1, 6, 2# hours, and 19 days. Instead of a monotonic decreas- ing relationship between retention and time, Khmin found a significant curvilinear relationship, with retention declin- ing significantly from O to 1 hour, then rising significantly from 1 hour to 19 days. To account for this relationship, two independent processes were postulated. The first segment of the curve extending from O to 1 hour was described in terms of a continued dissipation of positive transfer and agrees with the established vernacular concept of forgetting. Fer the rising segment of the curve extending from 1 hour to 19 days, Khmin suggested an incubation or Jelling of the avoidance response which begins to have a facilitory effect on avoidance performance about 1 hour after the initial learn- ing trials. In a replication of Khmin's study, Denny (1958) ob- tained similar results but offered a different interpretation. Denny explained the decline in retention of the 0-1 hour seg- ment of the curve by hypothesizing that anxiety builds up after the initial set of learning trials to the point of interfering with avoidance performance. This incubated anxiety reaches a peak in about 1 hour and then begins to dissipate. The rising segment of the curve, from 1 hour to 19 days, is explained in terms of the dissipation of anxiety. Experhmental evidence supporting this hypothesis was derived from Denny's study, in which two conditions designed to counteract the incubation of anxiety during the 1 hour inter- ..1 yielded results consistent with expectations. The present study represents an attempt to assess Denny's interpretation of the Khmin effect by using a physio- logical index of anxiety. More specifically, the amount of free HCl secreted in the stomach will be used as a measure of anxiety. The relationship between anxiety and free 301 secretion has led to considerable disagreement among investigators in the past. This seems to be due, in part, to a failure to recognize the differential reaction of gastric secretions to chronic anxiety as opposed to acute anxiety. Studies involv- ing chronic anxiety have generally revealed a positive re- lationship between anxiety and free H01. Mahl (1949) paired shock and buzzer over a prolonged period of time to develop a state of chronic anxiety in dogs. Significant increases in free HCl accompanied increases in anxiety. Here recently, a number of investigators (e.g., Sawrey (1956), Porter (1958), and Sines (1959)) have successfully produced gastric ulcers (primarily as a result of increased free HCl secretion) in animals under various forms of prolonged anxiety states. In direct contrast, the majority of studies employing acute anxiety have yielded a negative relationship between free H01 secretion and anxiety. Observations by Beaumont (183k) and Helf and Wblff (1943) on human subjects with gastric fistulas have revealed inhibiting effects of acute fear and anxiety on free HCl secretions. More recently, Mahl (1952) reported the results of a study in which three monkeys were subjected to two different buzzer-shock conditions de- signed to produce acute and chronic anxiety. Only in the chronic anxiety situation was free H01 found to increase significantly. A short preliminary study was conducted by the present author to further clarify the relationship of acute anxiety to gastric secretions. Three pylorus operated rats were given 25 avoidance trials in a shuttlebox. Mea- surement of gastric secretions after 1 hour showed a sub- stantial decrease in the volume of gastric secretions as com- pared to a control group. The data from some of the above studies are admittedly based on small numbers of observations. However,tthe evidence seems to indicate that hypersecretion of free H01 is character- istic of chronic anxiety states, and hyposecretion of free HCl is characteristic of acute anxiety states. Anxiety aroused during the 1 hour period in which Kamin (1957) and Denny (1958) found decreasing performance will be considered as ”acute.” In line with Denny's hypo- thesis, the incubation of acute anxiety during the 1 hour interval (following the initial set of learning trials) should result in an inhibition of free H01 secretion. Further, the hypothesized dissipation of anxiety following the 1 hour interval should be paralleled by a corresponding increase in free 301. METHOD The‘gs were 40 experimentally naive albino rats se- lected from the colony maintained by the Department of Psychology at Michigan State University. Twenty males and twenty females were used; they ranged in age from 90-100 days. Two female‘Ss died during the course of the experiment and were replaced with similar‘Ss. The colony was maintained on an 5g,;ib_feeding schedule and was housed in cages in a room adjoining the experimental room with #-5‘§s of the same sex in each cage. The‘fis were equated for weight, age, and sex, and with these restrictions were assigned at random to one of four groups of 10‘§s each. Groups A, B, and 0 comprised the experimental groups, group D served as the control to establish a basal level of H01 secretion. Four‘fis, one from each group, were run each day and this was continued for 10 consecutive days. W The apparatus was a modified Mower-Miller shuttlebox. The shuttlebox, 36 inches long, 4 inches wide, and 14 inches high, was of wood construction and painted flat black. For observation purposes, the entire front of the box was clear glass. The floor was constructed of 1/8 inch brass rods spaced approximately 1/4 inch apart. The experimenter could administer the unconditioned stimulus (Us), an electric charge through the brass rods, independently to either half of the grid floor. A 0. J. Applegate stimulator was used to deliver the continuous 1.7 milliamp charge to the grid. A.7O decibel buzzer attached to one side of the shuttlebox and activated by four dry cells served as the conditioned stimulus (OS). The onset and termination of the CS and US were controlled by the experimenter who manually operated individual switches. The treatments for the various groups followed the de- sign illustrated in Table 1. Avoidance learning of the type used by Kamin (1957) was the experimental task. The experi- mental groups A, B, and 0 were given 25 avoidance learning trials in the shuttlebox under the following conditions. The gs were placed in the apparatus and allowed a one minute adaption period followed by 25 avoidance trials with a one minute intertrial interval. The CS-US interval was 5 seconds and the CS was response terminated. Each trial ended either in escape or avoidance responses. Fellowing the initial set of trials, groups A, B, and 0 received an additional set of 25 trials at intervals of 1 hour, 6 hours, or 24 hours, respectively. Group D did not receive any avoidance learning trials. During both conditioning phases of the experiment ‘gs in this group remained in their cages. The experimental treatments of group D therefore differed from the experi- mental treatments of groups A, B, and 0 by the absence of avoidance trials, contact with the apparatus, and handling during the avoidance learning procedure. For all‘fis in the experiment the surgical operation of pylorus ligation, described by Shay (19A5), was performed. The Operation which required approximately 10 minutes was so timed that its completion proceeded the second set of avoidance trials by one hour. The experimental groups A, B, and 0 were differentiated in terms of the number of hours which elapsed between the end of the first conditioning session and the be— ginning of the second conditioning session. In group A,‘§s were pylorus ligated immediately after the initial set Of avoidance trials. One hour after the completion Of the oper~ ation Se in this group were returned to the apparatus for the second set of avoidance trials. The gm in group B were pylorus ligated 5 hours after the initial set of avoidance trials and were returned to the apparatus one hour later for the second set Of avoidance trials. The‘gs in group 0 were pylorus ligated 23 hours after the initial set of avoidance trials and were also returned to the apparatus for the second set of avoidance trials one hour after the completion of the Operation. In group D‘fis were pylorus ligated shortly after being removed from their cages. Since the operation of pylorus ligated was performed as a means to retain gastric secretions for analyses, certain precautions were necessary to eliminate contamination of the fluids. Contamination from undigested food was reduced by maintaining all‘Ss in the experiment on a 72 hour total food deprivation prior to pylorus ligation. Copraphagy was controlled during the deprivation period by housing the‘gs in cages of the same dimensions as their home cages, but without bedding and with raised 1/2 inch wire mesh floors. Post- Operatively all‘gs were placed in individual cages measuring 1 foot x 1 foot x 1 foot with raised 1/2 inch wire mesh floors. Immediately after the second set of trials, or in the case of group D, 1 hour and 26 minutes after pylorus ligation, all gs were sacrificed by over-anesthetistizing with ether. Death from Over-anesthetistizing took place within 30-120 seconds. The incision was reopened, and a hemostat was clamped on the esophagus near the diaphrmm. The stomach was completely excised from the body cavity and its contents drained into a 10 cc. graduated cylinder by releasing the hemostat. The gastric contents of each §_were analysed for determination of total volume, acidity (pH), percentage of free H01, percentage of combined H01, and percentage of total H01. The one drop sample electrode attachment for the "Beck- man Model H" pH meter was used for pH determinations. The test for free H01 was carried out using TOpfers Reagent as an indicator and titrating with 0.1 N. NaOH. A 1% alcoholic solution Of phenolpethalein was used as an indicator and titration again carried out with 0.1 N. NaOH in the test for total H01. Combined H01 was derived by subtracting the calcu- lated free H01 from the calculated total H01. RESULTS Behayigrgl Qgtg The mean number of avoidances and standard deviations for the three experimental groups, A, B, and 0, on original and relearning avoidance trials are presented in Table 2. The random assignment of‘és to groups to each of ten levels, equated for weight, age, and sex lent itself to a treatment x.levels analysis of variance design with one ob- servation per cell, as described in Linquist (1956). The between groups‘g for the number of avoidances during the original learning trials was 0.33 (2 and 18 df). This value falling short of significance indicates that initially the expermmental groups which had not yet been experimentally differentiated did not differ significantly on original avoidance learning. The between groupslg for the number Of avoidances during the relearning trials was 0.91 which is also far short of significance (2 and 18 df). Since the analysis of variance does not take into con- sideration the correlation between original and relearning scores, a double classification analysis of covariance as described in Edwards (1950) was used as a more precise test. The analysis is summarized in Table 3. A within groups correlation of 0.78 was found between the avoidance scores on original and relearning trials. The between groups‘z for the 10 11 adjusted means relearning scores was 3.93 (2 and 17 at). This value was found to be significant at the .05 level. Table 4 presents both the raw and adjusted mean scores for avoidances during relearning. Application of individual‘t tests to the three pairs of adjusted means yielded one significant difference. Group A (1 hour interval) made significantly fewer avoidances on the relearning trials than Group 0 (24 hour interval). (t = 2.7 p < .05 17 at) * Ph ica Table 4 presents the means and standard deviations of the physiological measure of total volume, free H01, combined H01, total H01, and acidity (pH) for each of the experimental groups and the control group. Free H01, combined H01, and total H01 are expressed in degrees which are units of acid concentration. The measure of free H01 was submitted to a treatments x levels analysis of variance. The between groups‘g was .09 (3 and 27 df) falling short of significance at the .05 level. The total volume of gastric secretion was similarly submitted to a treatments X levels analysis of variance. The between groups I was 2.78 (3 and 27 df), also falling short of significance at the .05 level. Although the primary interest in this study was the quantity of free H01 and the resultant quantity of total 12 gastric secretions as a measure of anxiety, all of the re- maining related physiological measures were analyzed with a treatments X levels analysis of variance. None of the between groups‘gs were found to be significant, and as Ob- served from Table 5, no strong trends among the group means were evident. DISCUSSION The results Of the behavioral data were consistent with the results from studies performed under similar condi- tions. Both Kamin (1957) and Denny (1958) found that the one hour interval group (Group A) made significantly fewer avoidance responses than the 24 hour interval group (Group 0) on the relearning trials. The only major discrepancy between the data reported by Khmin and the data of the present study is that better performance was Obtained on original learning and relearning by the present investigator. Kamin obtained a grand mean number of avoidances during original learning of 5.7 versus 9.3 for the present study; Kemin obtained a grand mean number of avoidances during relearning of 6.6 vermus 14.5 for the present study. The discrepancies in avoidance performance levels may be accounted for in terms of the results of a recent study by Thomas (1960). In this study, Thomas found that avoidance performance on original and relearning trials was considerably depressed in shuttleboxes with low ceiling heights. Since the shuttlebox used by Hhmin had a ceiling height of 4-3/4 inches as compared to a ceiling height of 14 inches in the present study, the apparent discrepancies in avoidance per- formance seems to be resolved. 13 14 The fact that the behavioral data yielded no other marked deviations from that of the previous investigations has particular significance in this study. The Se in this experiment were subjected to several unique conditions not present in previous avoidance learning studies. In the present study all §s were maintained on 72 hour food deprivation, all gs were anesthetized and pylorus ligated, and during the course Of the expermment there were two cage changes for all'gs. Although the possibility of isolating the individual effects of each of these variables is not possible in this study, the combined effects of food deprivation, pylorus ligation, and cage changes seams to have had little bearing on the outcome of avoidance learning and relearning behavior as compared with the findings of previous studies. The results of the analysis of physiological measures of total volume Of gastric secretion, free H01, combined H01, total H01, and pH failed to yield any significant differences among groups. However, the means of free H01 and total volume of gastric contents can be seen, from Table 5, to be in approximate alignment with the predictions from Denny's ”incubation of anxiety" hypothesis. That is, the one hour interval group (Group A) was considered to be in a state of acute anxiety and, therefore, was expected to have had the lowest output of free H01. The hypothesized dissipation of anxiety following the one hour interval was predicted to 15 be paralleled by an increase in free H01 in Group B (6 hour interval) and Group 0 (24 hour interval). The control group, having had no experience with avoidance learning and the re- sultant anxiety, was expected to yield the highest output of free H01. Since the total volume of gastric secretions would be affected by free H01 secretions, the same relationships were expected to hold for the total volume of gastric secretions. As seen from Table 5, the only case in which the Observed trends did not bear out these relationships was the finding of slightly more free H01 in Group A.(1 hour interval) than in Group B (6 hour interval). Although the trends of the means of free H01 and total volume of gastric secretions are in support of the ”incubation of anxiety" hypothesis, the failure to obtain statistically significant results leads to the suggestion of some possible influencing factors. A.primary factor responsible for the non-significant differences may have been that the anxiety levels of the four groups did not differ sufficiently for a significant difference in free H01 to arise. That is, free H01 as a.mea~ sure of anxiety may have been too insensitive under the condi- tions of this experiment. In addition, the effects of the 72 hour food deprivation, anesthetization, pylorus ligation, and cage changes, may have tended to further reduce the group differences in anxiety by bringing about an emotional leveling effect in all fie. 16 Finally, the relatively small number of Se employed may have considerably decreased the required level of pre- cision for this experiment. SUMMARY Previous studies on avoidance learning and relearning have reported a decrement in avoidance performance following a one hour interval between avoidance learning trials and re- learning trials. Denny hypothesized that the decrement in performance was due to the incubation of anxiety during the one hour interval. Improved avoidance performance following longer intervals between the two sets of trials would there- fore indicate a dissipation of anxiety. This study attempts to test Denny's "incubation of anxiety” hypothesis by employing free H01 secretion in the stomach as a measure of anxiety. The predictions were that the hypothesized incubation of anxiety during the one hour interval would result in an inhibition of H01 secretion. The hypothesized dissipation of anxiety following the one hour interval would be paralleled by an increase in H01 secretion. Three experimental groups of rats were given 25 avoidance learning trials in a shuttlebox followed by a second set of 25 relearning trials at intervals of 1 hour, 6 hours, and 24 hours, respectively. As a means of collecting gastric secretions, the surgical operation Of pylorus ligation was performed on all experimental Se 1 hour prior to the re- learning trials. A control group, which did not receive any 17 18 avoidance learning trials, was pylorus ligated for the purpose of establishing a basal level of H01 secretion. Immediately after the relearning trials or in the case of the control group, 1 hour and 26 minutes after pylorus ligation, all gs were sacrificed and the gastric secretions were collected for gastric analysis. Determinations of free H01, combined H01, total H01, total volume, and pH were made for each gastric sample. The results of the behavioral data were found to be consistent with the results of previous studies which were run under similar conditions. The trends of the means of free H01 and total volume of gastric secretion were found to be in approximate alignment with predictions. The results of the statistical analysis, however, yielded no significant differences. Factors such as the insensitivity of free H01 as a measure of anxiety and the influence of certain experi- mental treatments On the anxiety levels were discussed as possible reasons for the failure to yield significant results. 19 TABLE 1 Experimental Procedure and Treatments for the Experimental Groups and Control Group Experimental Treatment Group A Group B Group 0 Group 0 Initial set Of 25 avoidance learning trials yes yes yes rest Interval between termination of avoidance learning trials and Oper- 0 hour 5 hours 23 rest ation hours Pylorus Ligation yes yes yes yes Interval between end of Operation and on- set of 25 relearning 1 hour 1 hour 1 hour 1 hour trials rest 25 relearning yes yes yes 26 avoidance trials minutes rest Sacrifice with ether yes yes yes yes 20 TABLE 2 Means and Standard Deviations (SD) of the Number of Avoidances on Original and Relearning Trials for the Experimental Groups Original Learning Group Measure Tri als Relearning Trials Mean 10 . 2 1 2. 7 A SD 4.52 4.57 Mean 9.4 15 B SD 5.52 5.83 Mean 8.2 15.9 C SD 6.23 5.9 21 scene mo. es eeeeacaswaee mm.ne .n.mm m me.m_. ease: eeeesaeq m. mm.mam m.¢.~ mm.mme b>.mmm ow Assam seam mucosa mccspem m._wm o>.nmm am.mmm mm fleece ._.m_ a. amm eqaem enqmwe me.nen .ma senate eased: deseaecm m.me. ¢¢.en_ u...mm m uaopoq scespem m.em ..—no pm.om m mucosa mccspom a shaman we cpesaeum NM .w MN w «kw he soaps «as, use: no seesaw no coasom HO um masoam Heaacsascaum one you moaoom ceaseaoad one no eemdaaseoo no edmhdcn< m Hands 22 TABLE 4 Observed and Adjusted Mean Scores for the Number of Avoidances on Relearning Trials Group Observed Mean Adjusted Mean A 12.7 11.89 15 14.88 0 15.9 16.81 23 TABLE 5 Means of the Physiological Measures for the Experimental Groups and Control Group Volume Free H01 Comb. H01 Total H01 Group in co in degrees in degrees in degrees pH A 1.36 29.15 49.20 78.35 1.63 B 1.81 28.50 54.80 83.30 2.25 C 1.96 30.85 51.35 82.20 1.76 D 2.35 34.20 48.26 82.46 1.87 TABLE 6 Standard Deviations Of the PhysiolOgical Measures for the Experimental Groups and the Control Group VOlume Free 301 Comb. HCl Total H01 Group in cc in degrees in degrees in degrees pH A 0.63 26.43 68.21 33.81 0.78 B 1.09 34.15 21.47 23.83 1.03 C 1.92 23.32 12.88 20.04 0.62 D .88 28.03 12.41 19.62 0.84 APPENDI CES 25 TABLE 7 Number of Avoidances on Original and Relearning Trials for the Experimental Groups Group A Subjects N0. of avoid- 1 2 3 4 5 6 Z 8 9 10 Totals ances on original 10 7 14 14 3 18 13 9 6 8 102 learning trials NO. of avoid- ances on re- learning trials 9 12 17 12 7 20 17 6 15 12 107 Group B Subjects NO. Of avoid- 11 2 3 4 5 § 1 8 2 10? Totals ances on original learning trials 8 5 9 14 14 4 19 4 14 3 94 No. of avoid- ances on re- learning trials 12 10 15 17 23 4 22 18 18 11 150 Group 0 Subjects N0. of avoid- 1 2 3 4 5 6 10 Totals ances on original learning trials 8 1 12 15 5 7 0 19 12 3 82 NO. of avoid- ances on re- learning trials 15 12 13 2O 18 19 4 23 23 12 159 26 an m.nm m.mom v.0. mmm —.m— m.—mm m.n— masses 1%. 1w. IS: .3. a... an: 191 Hal 2 0 ¢.— 0 No. 0 m. m# W.— m 0 won #N new ON n.m N— m. m mm N 0 #ep #0 #.N w— m.— N mm N.N m.Np ¢.— 0 pop 0 New 0 ow n.n his n.n O— N; om —.N m mm m mm m e. #N b . — o 0 ¢ mm m on , m .m 8. 4 mm m; n O N.— ON #eN —N m. nemn m. N m0 m.n on m.— mm w.— mm New — wamwwm QOHHMMONM flwwmwwm QOHFMMOMN wamwm . “OH“?MMOMM wamm“ “OHPMMOMH C” no essao> no cmsao> no esdHO> o easHO> 114138.] c all 1.134311 Jewell macaw Hoaamoo use museum Hsomesdaeawm on» new Hum seam use umoaaoaoem enhance no essHO> Hausa m quda REFERENCES Beaumont, W. Experiments and Observatigns on the Gastric {pics and Physiology 0; Digestion. Boston: Lilly Wait 00., 1834. Denny, M. R. The "Kamin Effect" in Avoidance Conditioning. Ameg, Psychol., 1958, 11, 419. Edwards, A. L. e ime tal s n in Pa cho o ical Rgeeaych. New York: Rinehart and Company, Inc., 1950. Kamin, L. J. The Retention Of an Incompletely Learned Avoidance Response. ’ C sio ho , 1957. 52, 457-460 e Lindquist, E. F. Design and Analysig O; Eypeyimegts in Psychology and Education. Boston: Houghton Mifflin, 195 . Mahl, G. F. The Effect of Chronic Fear on the Gastric Secre- tion: Of H01 in Dogs. Ps chosomatic Med , 1949, ll, 30-4 . Mahl, G. F. Relationship between Acute and Chronic Fear and the Gastric Acidity and Blood Sugar Levels in Macaca Monkeys. Psyghosomatic Med,, 1952, 14, 182-210. Porter, R. V., Brady, J. V., Conrad, D., Mason, J. V., and Galambos, R. Some Experimental Observations on Gastrointestinal Lesions in Behaviorally Conditioned MOnkeys. ngghosomatic Med,, 1958, gg, 379-394. Sawrey, W. L. An Experimental Method of Producing Gastric Ulcers. g, Comp. Physiol, ngchol,, 1956. 4 . 269-270e Shay, H. A Simple Method for the Uniform PrOduction of Gastric Ulceration in the Rat. Ga 9 O , 1945, 5, 43-6‘e Sines, J. O. Selective Breeding for Development of Stomach Lesions Following Stress in the Rat. J, QQmE‘ Physiol, Psychol,, 1959, 2, 615. Thomas, J. O. Avoidance Learning and Relearning as a Function of Shuttlebox Dimensions. Science, 1960, 132, 620-621. WOlf, S., and WOlff, H. G. Human Gagtrig Eungtiog. New IOrk: Oxford University Press, 19 3. 27 MHMIGMJ STATE UNIVFRSIH uswsmgq ‘9 ': '4! 1:1] 5“! l «1111:1111» ‘ 3 1293' 03 ‘ 175 099 1 ‘ Hi ““ 11‘ 5 ”ill ll 1! 9