A.N. Mulla, J.
1. An armed dacoity was committed at the house of one Ramratan, a resident of village Bechepurwa, police station Kotwali, Lakhimpur, district Kheri, on the night between the 19th and the 20th of February, 1953. Ramratan resided with his brothers Bahuram and Ram Autar and other members of the family in separate portions of the same house. At about midnight Ramratan was disturbed in his sleep when a dacoit jumped inside the house and unchained the door. Immediately afterwards, 10 or 12 more dacoits entered the house through the main door. These dacoits were armed with pistols, guns, spears and lathis. Seeing the dacoits, Ramratan shouted and his brothers rushed into the courtyard. One of the dacoits fired a shot which hit Baburam. The dacoits then started plundering and this gave an opportunity to Ramratan to slip out of the house. He raised an alarm and, as a result, the neighbours collected and some one set fire to the cattle-shed of Ramratan.
During the course of the dacoity the residents of village Mathurapurwa, which is at a distance of about 3 or 4 furlongs from Bechepurwa, also arrived The burning cattle-shed threw a bright light and the villagers outside saw the features of the dacoits in this light. When the dacoits came out after collecting the looted property, they were pursued by the villagers and then the dacoits fired some shots which injured some villagers including Kumari Maikin. The dacoits could not be pursued after that and they escaped.
2. After committing dacoity at Bechepurwa the dacoits went to village Mathurapurwa and committed another dacoity at the house of Khushi Ram, a resident of that village. In this dacoity also there was an encounter between the dacoits and the villagers. Ramratan as well as other residents of Bechepurwa, when they came to know that a dacoity was being committed at Mathurapurwa, went there. In the encounter that ensued two persons named Inder Bahadur and Chakkardinlost their lives. None of the dacoits could, however, be apprehended and they escaped.
3. Next morning, Ramratan lodged the first information report at police station Kotwali, Lakhimpur, at 10.45 A.M. Lakhimpur is about six miles from village Bechcpurwa. Ramratan had gone on a bullock curt with all the injured persons when he lodged the report. No report was lodged bv Khushi Ram separately and it was Ramratan who lodged a comprehensive report in which the facts of both the dacoities were mentioned. When the encounter had taken place at Mathurapurwa, the dacoits had left behind some articles and they were taken by Ramratan to police station Kotwali when he wont there. Several used cartridges, two fountain pens, one old chadar and a lathi were the articles which were deposited by Ramratan when he lodged the report.
4. The investigation of these two crimes which were recorded as one was conducted by Sri Raj Kumar Singh, the second officer of police station Kotwali, Lakhimpur, He first went to Mathurapurwa and there prepared the inquest report of the dead body of Inder Bahadur. Chakkar who was still alive was sent to hospital. After conducting some preliminary investigation at Mathurapurwa, he came to Bechepurwa and prepared site plans of both the scenes of the dacoities. Thereafter he examined witnesses and sent the injured persons to Lakhimpur for medical examination of their injuries. The dead body of Inder Bahadur was also sent to Lakhimpur mortuary for post mortem examination.
5. As a result of his investigations,' the investigating officer came to the conclusion that the same gang of dacoits had committed both the dacoities. The chadar that was left behind by the dacoits was shown by him to persons at Baragaon, a big village, which is four miles away from the scene of the dacoities. This chadar was identified by sonic persons as that of Anwar appellant. Ho thereupon arrested Anwar appellant on the 4th of March, 1953. Faqirey, another appellant in this case, was arrested on the 3rd of April, 1953.
6. While this investigation was going on, it is alleged that two encounters took place on the night between the 29th and the 30th of March 1953 at two different villages. These villages are Patrasi and Ramnagri, In these encounters the villagers of these two places succeeded in injuring and arresting some dacoits. Two of the accused in this case were arrested in the Patrasi encounter and Babu alias Akbar appellant was injured and arrested in the Ramnagri encounter. As Babu was injured, he was sent to Lakhimpur Hospital where he was kept in a closed room and then he was sent to Lakhimpur jail with his face covered. Another accused who was arrested in connection with these two dacoities was one Balram Das. This arrest took place on the 1-4-1953. It is not necessary to mention the arrests of the other suspects in this case.
7. After the suspects were arrested, they were put up for identification. As they were arrested on different dates, several identification parades were held. Anwar and Faqirey appellants were put up for identification on the 9-4-1953, and this parade may be called the first parade. The remainingaccused including Babu appellant were put up for identification on the 1-5-1953. This may be called the second parade. The investigating officer Sri. Raj Kumar Singh, who was present outside the jail when this parade was conducted, gave a report on that date that he did not want Babu to be placed for identification before any other witness.
Babu could be identified by one witness only in this parade and as this did not suit Sri Raj Kumar Singh, ho went back on this report and prayed that Babu should be put up for identification before some other witnesses again. This parade was held on the 16-5-1953 and this may be called the third parade. Balram Das who was arrested on the 1-4-1953 was put up for identification in a separate parade held on the 1-8-1953 and this may be called the fourth parade. As the results of these parades were satisfactory the suspects were prosecuted. In all seven persons including the three appellants and Balram Das were prosecuted in this case.
8. The defence of the appellants was that they were known to the witnesses from before and they were also shown to them after arrest. Balram Das contended that he was falsely implicated because of his enmity with the police. Babu appellant alleged that he had a dispute with one Bachchu-singh who had fired at him and he was then handed over to the police at Lakhimpur Kotwali and was implicated in this case. He also alleged that ha was shown, to the witnesses at Lakhimpur Kotwali. The defence of the other suspects also was similar. Anwar and Faqirey appellants examined a witness each to establish that the prosecution witnesses knew them from before, but this evidence is worthless and can safely be ignored.
9. It would be seen from the history of the case given above that the evidence against the accused in this case consisted of the evidence of identification only. The trial, court accepted this evidence in the case of the three appellants, but rejected it in the case of the other suspects. It, therefore, convicted Anwar, Faqirey and Babu under Section 395 I. P. Code and sentenced each of them to four years rigorous imprisonment each. We may observe here that if the trial court came to the conclusion that these three appellants had committed this armed dacoity, the sentence inflicted erred on the side of leniency.
Perhaps the trial court did so, because at a very late stage when the case was ripe for arguments it was contended before it that these two dacoities cannot be tried together and it split up the trial and tiled the offenders in two different cases. As the trial court gave a sentence of life imprisonment to Anwar and Faqirey appellants in the Mathurapurwa dacoity case, perhaps it was for this reason, that a light sentence was inflicted in the Bechepurwa dacoity case. The trial court acquitted Babu in the Mathurapurwa dacoity case.
10. It has not been contended before us that a dacoity was not committed at Bechepurwa. The counsel, however, urged that there is no evidence that the same gang of dacoits committed both the dacoities and he relied upon the observations of the trial court in support of this contention. The trial court has given some reasons in its decision forcoming to the conclusion that the prosecution hasfailed to establish that the same gang of dacoits committed these two dacoities. In our opinion the trial court ignored the dominant circumstance namely that the two dacoities were committed in such close proximity of time and place. This circumstance in our opinion, over-rides the reasons which have been advanced by the trial court. We are of the opinion that there can be hardly any doubt that it was the same gang of dacoits who committed these two dacoities.
11. This question is, however, quite unimportant. The real question is whether the identification by the residents of one village can be used for proving that the dacoits took part in the dacoity which was committed in the other village. As the two trials have been separated, it is not necessary to disagree with the view taken by the learned trial court on this point. The learned trial court came to the conclusion that the residents of Bechepurwa did not go to Mathurapurwa and similarly the residents of Mathurapurwa were not competent witnesses of the Bechepurwa dacoity. It, therefore, relied upon the identification of only those witnesses who were the residents of the village concerned.
12. There are certain aspects of the case which indicate that the investigation was not straightforward. We would first take up the first information report. The trial court did not specifically come to this finding but it can be inferred from its observations that it came to the conclusion that in order to minimise the number of dacoities the investigating agency treated these two dacoities as one dacoity and incorporated the facts of both the offences in one report. In our opinion the trial court was justified in coming to this conclusion. The report, therefore, ceases to be a spontaneous statement made by the victims of the dacoity.
It becomes a dressed up version. Ramratan disowned all responsibility of giving any details of the Mathurapurwa dacoity. Khushi Ram, on the other hand, stated that he dictated no part of the first information report. It is, therefore obvious that it was the investigating agency who as a result of its preliminary investigations incorporated the facts of the two crimes in one report and gave Ramratan the role of dictating this report. The length and the details contained in this report also make it highly probable that it was not the unaided effort of Ramratan. If Ramratan could marshal so many details it is surprising that he could not describe the dacoits in a better manner. The only description of the dacoits given in the first information report is as follows:
'Some of the dacoits were wearing under-wears; while some were wearing dhotis. They were speaking in country dialect and were accosting each other as Babuji.'
This performance is in marked contrast with the details of the incidents of two separate diacoities which are alleged to have been orally given by Ramratan. We have specially drawn attention to the description of the dacoits, for at a later stage we would use it to show that the investigation was not honest, but tainted.
13. Admittedly there was plenty of light in both these dacoities. In the Bechepurwa dacoitywhen the thatch of Ramratan was lighted the fire spread and several other houses were burnt. Even some cattle which were inside the cattle-shed were burnt. It, therefore, cannot be said that the light was not sufficient fox the witnesses to see the features of the dacoits. In addition to this the story of a pursuit and an encounter with injuries on the persons of several villagers show that they came in close contact with the dacoits. Similarly in Mathurapurwa dacoity also thatches were burnt and an encounter took place with the dacoits.
It was even alleged that lathis and swords were used by the villagers and some of the dacoits were injured. Under these circumstances it is extremely surprising that the witnesses who were present in a large number when the first information report was lodged by Ramratan could give no better description of the dacoits than mentioned above. Such a description could only have been given if the dacoits possessed no distinctive features or if they had taken precautions to cover their distinctive features. It is not possible to accept that Ramratam was so dumb that though he could dictate such a detailed first information report, he could not give a better description of the dacoits.
14. When the witnesses were interrogated by the investigating agency, they again failed to give any description of the dacoits. Obviously the investigating agency must have interrogated these witnesses with a view to trace out the offenders, and they must have questioned the witnesses about the features of the dacoits. If in spite of such a questioning the witnesses could not give any description, it only presents two possibilities (i) The dacoits were not well seen and, therefore, the witnesses could not give any description. In view of the light that existed in this case and the encounters that took place between the witnesses and the dacoits, it is difficult to hold that the light was not sufficient. If we come to this finding in such a case, it would be difficult to find in any case that there was sufficient light.
We are, therefore, satisfied that this possibility is not applicable to the circumstances of the case. (2) The investigating agency made no attempt to find out the description of the dacoits from the witnesses as it intended to prosecute its own list of suspects. This seems to us to be the probable explanation of this lack of description both in the first information report and the statements made by the witnesses during investigation. This suspicion is greatly strengthened by the prosecution of Balram Das in this case. The trial court acquitted Balram Das. No less than 9 persons had identified him in the jail parade. The trial court observed:
'Balram Das is easily 6 feet in height. His long hair of the head are flowing upto his shoulders. He has also a big beard. Apart from the height, the beard and the long hair on the head of Balram Das are such features that no one could fail to notice them if Balram Das had been among the dacoits. It is not mentioned in the F. I. R. Ex. PI that any dacoit had a long beard and long hair on the head. No one mentioned to the investigating officer during the course of the investigation that one dacoit had a long beard and long hair on the head. The witnesses who had identified Balram Das couldgive no reason why these special features of Balram Das were not disclosed to the investigating officer. In my opinion the absence of prominent features of Balram Das in the F. I. R. and the absence of the description of the features of Balram Das in the statements of the witnesses made to the investigating officer go to show that no dacoit of the description of Balram Das was among the dacoits.'
In our opinion no exception can be taken to the reasoning of the learned trial court on this point. It is, therefore, clear that the investigating agency wanted to implicate Balram Das in this case.
15. There is another circumstance which makes this extremely probable. We have mentioned above that Balram Das was arrested on the 1-4-1953. Balram Das was not put up in an identification parade in connection with this dacoity upto the 1-8-1953. No explanation has been offered why Balram Das was not put up in an identification parade for four, months. Balram Das, it appears was primarily arrested in connection with some other offence and the investigating agency for its own reasons decided to prosecute him in this case, also.
The way Balram Das was identified again strongly points to his being shown to the prosecution witnesses after arrest. Balram Das was put up in a parade on the 1st of August and two dacoities had occurred on the night between the 19th and the 20-2-1953. The parade was, therefore, held about 51/2 months after the date of the dacoity. One would expect that after such a long lapse of time the majority of witnesses would not be able to identify him as the image in their mind would fade out, but when this parade was held no less than 9 witnesses succeeded in picking out Balram Das. This remarkable feat of memory and observation when it is committed by so many witnesses is by itself a circumstance to indicate that the identification is not a result of observation but of something else.
The trial court explained this by holding that Balram Das had distinctive features, which we have mentioned above, when he was put up in the parade. This explanation does not go far enough. A person is not identified on account of distinctive features but because the witnesses are told about these distinctive features. We are, therefore, satisfied that at any rate in the case of Balram Das the investigating agency gave the witnesses either an opportunity to see him or they described his distinctive features to these witnesses which enable so many of them to pick him out. This naturally discredits the evidence of identification to a great deal even in the case of other suspects.
16. The way Anwar appellant was traced also seems suspicious to us. According to Sri Raj Kumar Singh, he showed the old used Chadar left behind by the dacoits at Mathurapurwa to the residents of Baragaon and there P. W. 40 Shahzad and some others recognised this Chadar as that of Anwar. It is not possible for us to accept that an old used Chadar is an identifiable article. No distinctive marks existed in this Chadar and one old used Chadar would be very much like another old used Chadar. We are, therefore, of the opinion that whatever the sources of information might be theinvestigating agency did not reach Anwar by means of the old used Chadar that was left behind.
17. The conduct of Sri Raj Kumar Singh in not prosecuting a person who was named by Chakkar deceased in his dying declaration also makes the investigation highly tainted. Chakkar received a stab wound in his chest and abdomen and he died in the district Hospital, Kheri, on the 20-2-1953 at 6.30 P. M. As his condition was dangerous his dying declaration was recorded at the Hospital by a Magistrate. In this dying declaration Chakkar named one Dirgaj Singh of Lakhnapurwa as one of the dacoits. Sri Raj Kumar Singh not only failed to prosecute Dirgaj Singh, but suppressed this dying declaration. We will quote an extract from the statement of Sri Raj Kumar Singh himself:
'The Kotwal had told me that dying declaration of Chakkar had been recorded. Bhagwan Swarup, the court moharrir of the Court of Extra Magistrate had given me a copy of the dying declaration. That copy is attached to the case diary. I did not see the original dying declaration. I did not file the dying declaration as I did not think it necessary to file it. In the copy of the dying declaration, one dacoit was named. I did not challan that man and did not get him identified by any witness. That man lives at a distance of half a mile from Becheyman lives at a distance of half a mile from Bechepurwa. Dirgaj Singh of Lakhnapurwa wa,s named in the copy of the dying declaration.'
It is surprising that Sri Raj Kumar Singh succeeded in losing the original dying declaration and the trial court also made no effort to trace it. It was a most important piece of evidence and the prosecution was permitted to suppress it. The least inference that can be drawn is that this dying declaration, if produced, would have been unfavourable to the prosecution case, Sri Raj Kumar Singh by not prosecuting Dirgaj Singh destroyed the evidentiary value of identification himself. There is no indication that Chakkar had any animus -against Dirgaj Singh when he named him.
He had received an injury from front and by means of a sharp-edged weapon, which would bring his assailant close to him. With all the alleged light that existed, he was in a very good position to identify his assailant. His evidence alone could have been accepted by the trial court to convict Dirgaj Singh and yet Sri Raj Kumar Singh did not even prosecute him and we do not know whether Dirgaj Singh was even apprehended. These circumstances raise something more than a mere suspicion in our minds that Sri Raj Kumar Singh entered into a deal with Dirgaj Singh which was profitable to both sides. The real culprits were permitted to escape and the investigating officer was ready with his list of substitutes.
18. In assessing evidence of identification it is neither possible nor desirable to lay down any hard and fast rules, for there is an infinite variety of situations and circumstances in which a witness gets an opportunity of seeing the culprits, but still we are of the opinion that some general guiding principles can be stated. The trial courts are repeatedly making the mistake of following a rule of thumb when they assess the evidence of identification andthey do not subject it to the tests of prudence. It is one of the basic principles of criminal law that a fact or circumstance should be proved against an accused before, it can be relied upon and usedagainst him.
A fact is held to be proved only when it fulfils the definition of the word 'proved' given in Section 8 of the Indian Evidence Act. The evidence of identification is as much subject to this definition as any other kind of evidence, but it seems to us that in assessing the evidence of identification thetrial courts do not apply the tests provided in this definition. It is true that an absolute certainty is not needed, but the court has to test the evidence with prudence and accept, it only when it is so highly probable that its truth can safely be accepted. What is the approach of prudence? Prudence in our opinion requires that the court should approach the evidence with the reasonable doubts of an intelligent person and accept it only if those doubts are removed. This does not mean distrusting the evidence, but only subjecting it to reasonable tests. It certainlyexcludes from its orbit the blind faith of a true believer, for prudence and credulity do not go together. It is this second kind of approach which is coming to our notice frequently in dacoity cases.
19. What in essence is the evidence of identification? At its best it is the recognition of a stranger's face seen momentarily during the stress and excitement of an incident in torch light or some other artificial light after a lapse of several days or months, as the case may be, and the identifierclaims that he carried this image all this time in his mind and when he went to the identification parade the image in his mind corresponded with the features of a suspect in the parade whom heaccordingly picked out. It can, therefore, be accepted only if the court is satisfied on the following points.
1. The witness had at least a fair, if not a goodopportunity of seeing the dacoits. This naturally raises the subsidiary questions of sufficiency of light and the proximity of the witness to the offenders.
2. The identification parade was held within a reasonable time of the incident. The image of a stranger's face is likely to fade out as the days pass by and the court should be satisfied that when the witness picked out a suspect he did not do so after such a length of time that a person possessing a normal power of observation and memory was not likely to retain the image for the relevant duration.
3. The power of observation of the witness is reliable. This should be tested from two angles:
(a) That when the parade was held the witness was really tested and it was not made too easy for him to pick out the suspect. Questions like these arise. Were sufficient number of persons mixed with the suspect or not? Were proper precautions taken against his distinctive marks or not? Was the identification parade a genuine test or a farce?
(b) That the witness did not commit so many mistakes which would make a reasonable person doubt his capacity for observation. These mistakes whether committed in the same parade or in some other parade are to be considered in assessing thepower of observation of the witness. It may be that for satisfactory reasons the court may not attach, Importance to these mistakes, but they cannot be completely ignored.
4. The statement of the witness that he did not know the suspect from before is believable. Such circumstances for example that the accused lived in the adjoining village or on account of relationship or business he occasionally came to the village of the witness should be considered in connection with this point.
5. That the witnesses were not given an opportunity to see the accused after their arrest. The most important point in this connection is whether the investigation conducted in the case inspires confidence or not.
The moment the court finds that the investigation is tainted, it should be on its guard. While small delinquencies may not destroy the evidence of identification, but where there are major delinquencies and the investigation is highly tainted, it would be against the rule of prudence to accept the evidence of identification. Once the investigation becomes tainted the evidentiary value of the police records is gone and they cannot be relied upon. In such a case the fact that the investigating agency was hostile to an accused and this could have been a reason to implicate him falsely also becomes important.
20. We have given a sketchy outline of the considerations that should be weighed by the court before it makes its final assessment and in reaching its conclusions the court should always, follow the golden rule of prudence -- namely that the probable should be preferred to the possible, unless the possible is supported by irreproachable evidence.
21. The reason why the evidence of identification is to be subjected to these tests is quite obvious. The evidence of identification at its best is a weak type of proof for it depends entirely upon the capacity of a witness to register a true impression in his mind and then keep it in his memory. The chances of a mistake are far greater in this type of evidence than where the witness, deposes about facts within his knowledge.
An honest witness can seldom say more than this that the man whom he picked out in the parade resembled the man whom he saw at the time of the incident. The resemblance can so often be deceptive. The human face has so much in common that it is only the formation of details which puts a distinctive stamp upon the face of an individual. This truth is brought home to us when we see people of a different country or race Even for persons whose eyes are far more trained and observant than the eyes of ordinary villagers, it is not easy to pick out a Chinese of a Japanese whom, they have cursorily seen before, from amongst a group of their countrymen. They all look so much alike.
The reason is that the individual characteristics of a face do not get impressed upon the mind unless a man is frequently seen and only the broad general features are noticed. It is well known that a basic resemblance exists between the members of a family and acquaintances who do not know them intimately are frequently mistaking one member for another.
In those cases where a person disappears or is wanted by the Police and a detailed description of his appearance is published it is not easy to spot the individual with the help of this description for it seems to fit so many persons.
We have enumerated all these facts only to stress that the evidence of identification at its best is of a weak character and it can be accepted only where the investigation is above board. Where the investigation inspires confidence and the witness was fairly tested before he picked out the suspect the demands of prudence are satisfied and the court feeling reassured rejects the possibility of a coincidence, which becomes remote. But where the investigation is tainted the performance of the witness cannot be accepted at its face value for the reasonable possibility of external aid being given to the witness cannot be eliminated.
22. It was perhaps, for these and similar other reasons that the High Court has from time to time laid down certain rules of caution in order to test the observation and memory of identifying witnesses. The trial courts are however frequently either ignoring or misunderstanding and misapplying these rules. Only two of these rules are relevant to this case. We will confine our observations to these two rules.
23. The first rule relates to the number of under-trials to be mixed with a suspect in order to eliminate the reasonable possibility of a chance identification and to make the results of identification acceptable. The second rule stresses that the performance of the witnesses in other parades is also relevant in assessing his power of observation.
24. In Emperor v. Chhadammi Lal : AIR1936All373 a Bench of this Court made the following observations :
'As regards the evidence of the identifying witnesses the respondent was mixed up with only five under trials. The value of identification depends on the factor which minimize the possibility of chance as much as possible, It appears that' a practice has been established on account of certain observations made in the judgment in Asa Ram Gang case to mix 5 under-trials with a suspect for purposes of identification proceedings. The observations in the judgment seem to have been misunderstood because those observations were made in connection with a case in which there were a large number of suspects who were put up for identification. In cases in which there is only one or two suspects to he put for identification, the proportion of 1 to 5 cannot be regarded as satisfactory. There should be at least 10 under-trials for each suspect in such cases because every effort should be made to minimise the possibility of a chance which in the first instance can be done by mixing as many persons as possible with the suspect who is put up for identification.'
To the best of our knowledge this decision has neither been distinguished nor over-ruled by any subsequent decision of the High Court. In our opinion it lays down a salutary rule of caution and it was accepted and followed by us in State v. Bharosey, AIR 1955 NUC (AH) 5287. This rule was again reiterated by another Bench decision of this Court in State v. Supra, Criminal Appeal No. 41of 1955, Lucknow Bench, decided only last month on the 24th of April, 1957. Again a learned Judge in Satya Narain v. The State : AIR1953All385 , although he did not refer to Chhadammi Lal's case : AIR1936All373 , observed as follows at page 394;-
'The proper way to hold identification proceedings is to put up each suspect separately for identification mixed with as large a number of innocent men as possible, in any case not less than nine or ten.'
The reason for this rule is clear. It is an attempt to ensure that the identifying witnesses would be subjected to a real test and it would not be made too easy for them to pick out a suspect. The trial court accepted this rule of caution in the case of Balram Das but for unaccountable reasons it ignored this rule in the case of Anwar and Faqirey appellants. The trial court while discussing Balram-Das's case observed :
'Sri K. M. Ali (P.W. 27) who conducted the identification proceedings of Balram Das stated that he mixed 8 under-trials with Balram Das when Balram Das was put up for identification. In a case reported in : AIR1936All373 , it has been held that where only one or two suspects are put up for identification then at least ten under-trials should be mixed. As, however, only 8 under-trials were mixed with Balram Das so the evidence of identification against him cannot be accepted.'
25. Anwar and Faqirey appellants were put up for identification in the first parade held on 9-4-1953 and only 14 under-trials were mixed with them. There was thus a clear breach of the rule of caution mentioned above. It is surprising that the trial court considered the ratio of 8 to 1 unsatisfactory in the case of Balram Dass, but it felt no hesitation in accepting the evidence of identification against these two appellants, when the ratio was only 7 to. 1. Perhaps the trial court overlooked the fact that only 14 under-trials were mixed with these two appellants.
While we would not like to make a categori cal observation that the ratio 7 to 1 in the case of one or two suspects ipso facto destroys the results of identification we feel no hesitation in observing that it considerably diminishes the value of identification and unless the investigation is absolutely above board, it would not be prudent to place any reliance on such identification. We have already given our reasons above for holding that the investigation in this case was highly tainted, and so in our opinion it would not be safe to accept the results of identification in this case.
26. The learned counsel for the State has placed certain decisions before us in which a proportion of five under-trials to one suspect was considered satisfactory and in some cases even a lesser ratio, though adversely commented upon, was accepted. So far as the second category of cases are concerned, we may say so with respect to those Judges who gave those decisions that their view is unacceptable to us. Their view, in our opinion, is in conflict with the definition of the word 'proved' in Section 3 of the Indian Evidence Act. Unless a witness is fairly tested it is not possible to accepthis identification, and a ratio of less than 5 to 1 is not sufficient to test the witness as it makes it easy for him to pick out a suspect. It is, therefore, not necessary for us to enumerate these decisions.
27. As regards the other category of decisions in which a ratio of 5 to 1 was considered satisfactory, our attention was drawn specially to two Bench decisions of our High Court. These two cases are : (1) State v. Wahid Bux, AIR 1958 AH 314 and(2) Dal Chand v. The State : AIR1953All123 . The learned Judges in Wahid Bux's case : AIR1953All314 , observed :
'Of course, it is always better to have as large a number of persons mixed up with the accused as possible. But no hard and fast rule can be laid down and we are of the opinion that if five times the number of the accused persons are mixed with them, it cannot be said that there is any flaw in the identification proceedings.'
This point of view was acceptable to the learned Judges who decided Dal Chand's case : AIR1953All123 . They observed:
'Learned counsel was not able to point out to us any authority in support of the proposition that the proportion of 5 to 1 between the number of under-trials mixed and the number of suspects was defective in law. We are not aware of any such law, nor has the learned counsel been able to point out to us any such law.'
It would be seen from the extracts quoted above that Chhadammi Lal's case escaped the attention of the learned Judge in both these cases. Again there is no conflict between these decisions and the rule of caution laid down in Chhadammi Lal's case. One relates to identification proceedings where there are a large number of suspects put up in the parade and the other relates to proceedings where there are only one or two suspects. The trial courts have omitted to notice this difference and they seem to think that the two decisions cited above are also applicable to those proceedings where only one or two suspects are put up for identification. This error on the part of the trial court was noticed in Cri. A. No. 41 of 1955, D/- 24-4-1957 (All), and it was observed :
'Thirdly we find that the respondent and one other suspect were put up in a parade in which only 15 other under-trials were mixed. The rule of caution laid down by this Court in more than one Bench decision is that where there are one or two suspects, the ratio of under-trials to be mixed with the suspects should not be less than ten to one. It is, therefore, clear that the rule of caution laid down by the High Court was ignored by the Magistrate when he conducted the identification proceedings.
Either the Magistrate was ignorant of these decisions or he took his law not from the decisions of the High Court but from some other source. At any rate we see no reasons to ignore the salutary rule of caution laid down by the High Court and the results of the identification on this ground alone could have been discarded by the trial court. There seems to be a misunderstanding in the minds of the trial courts about the decisions of the High Court on this point. In some cases, where a largenumber of suspects were put up in the parade, the High Court has laid down that a ratio of five under trials to one suspect is not unsatisfactory.
Where the trial courts go astray, is that they..........apply the observations laid down in cases of a large number of suspects to cases in which there are only one or two suspects put up in the parade. There is no conflict in these decisions. Only the trial courts have failed to understand that a ratio of 5 to 1 may be acceptable in those parades where there is a large number of suspects, but a ratio of 10 to 1 is desirable where there are only one or two suspects. The reason is quite obvious. There should be two considerations before a Magistrate when he is conducting an identification parade.
The first consideration is that witnesses should be really tested and this necessitates that quite a large number of under-trials should be mixed with each suspect. The second consideration is that the parade should not become so unwieldy that witnesses may get bewildered by the numbers in the parade. It is by synchronising these two considerations that a correct approach can be made. Where there are only one or two suspects mixing of 10 under-trials with each suspect, will, not make a parade unwieldy, but if there are 7 or 8 suspects and the same rule is applied, then the parade will become unwieldy. It is for this reason that a ratio of 5 to 1 is acceptable in the case of a parade where there are several suspects, but where there are only one or two suspects, this ratio does not satisfy the rule of caution.'
We are in agreement with the observations made above. In our opinion a graduated scale of the number of under-trials to be mixed with the suspects would be most desirable. For example if there are three suspects in a parade, the ratio between the suspects and the under-trials may be reduced to 8 to 1 and this proportion may be decreased progressively when the number of suspects becomes more and more, but in no ease, whatever the number of suspects might be, it should become less than 5 to 1.
28. As in our opinion the investigation was tainted and the number of under-trials mixed in the parade was unsatisfactory, we do not think it safe to accept the evidence of identification against Anwar and Faqirey appellants.
29. Coming to the case of Babu appellant, we find that the trial court committed two glaring errors in assessing the evidence of identification against him. Firstly, it completely ignored the fact that this appellant without any explanation was put up in two parades. This circumstance by itself should have raised a doubt in the mind of a prudent person, but it was not even considered by the trial court. While it is open to a court to accept or reject a grave criticism that can be levelled against the testimony of a witness, it is certainly not open to it to close its curs and ignore the criticism altogether. It is the duty of a court to present a true picture of the case and where important criticisms are not even mentioned in the decisions the judgment ceases to be a judicial determination as it presents only an incomplete picture.
30. We have mentioned above that Babu was first put up for identification on 1-5-1953. On that day Sri Raj Kumar Singh submitted a report that he did not want Babu to be placed for identification before any other witnesses in any subsequent parade. Babu, however, could not be satisfactorily identified in this parade and thereupon Sri Raj Kumar Singh went back upon his report and called some other witnesses to a second parade which was held on the 16th of May, 1953. When Sri Raj Kumar Singh was cross-examined on this point, hecould give no reason why after giving up certainWitnesses he again insisted that a second parade should be held. Any prudent man would have hesitated to accept the results of the second parade,but the trial court considered this circumstance as of no importance and ignored it completely.
In our opinion the circumstance mentioned above lends strong support to the contention ofBabu that the witnesses were given an opportunity to see him. It is at any rate the most natural explanation of the conduct of the investigating officer. The trial court convicted Babu appellant uponthe identification of two witnesses P.W. 2 Chandrabahal and P.W. 10 Baldeo. Baldeo was not sent to the parade held on the 1st of May, 1953 and heidentified Babu on 16-5-1953. His identification is absolutely worthless and it must be discarded fromconsideration. This leaves only one identification against Babu and it cannot be considered sufficient to prove the case against him.
31. The second error committed by the trialcourt was that when assessing the evidence of identification it ignored the mistakes committed by someof the witnesses in other parades. The trial courtobserved:
'Babu has been correctly identified in jail identification, in the court of the committing Magistrate and in the Court of Session by Chandra Bhal P.W. 2 and Baldeo P.W. 11 None of these witnesses made any mistake in jail identification.'
This statement is again a half truth for basically it is incorrect. These witnesses did not make any mistake when they picked out Babu, but they mademistakes in the other parades. It is repeatedly coming to our notice that the trial courts have misunderstood the rule laid down by the High Court and they completely over-look the mistakes committed by the witnesses in the other parades. TheHigh Court to the best of our knowledge never laiddown that the mistakes committed by witnesses in other parades are wholly irrelevant. The decisionsof the High Court were explained in Bechu v. State, 1957 Cri. LJ 113 (All), by one of us and itwas observed :
'It is only those parades which are held long after the relevant parade which could be ignored from consideration. The parades which are held prior to the relevant parade cannot be ignored. Similarly parades held almost simultaneously or shortly after the relevant parade have also to be considered.'
We are in agreement with the view expressed above for any other interpretation would amount to a deletion of Section 146 of the Indian Evidence Act. It cannot be doubted that where a witness makes a mistake in identifying a suspect it discounts to a certain extent his power of observation and it isfor the court to decide whether the mistakes whichhe has committed are sufficient to make his identification doubtful or not. To hold that only the performance in the relevant parade should be considered is almost equivalent to holding that the circumstances which discredit a witness should not be considered.
The High Court never meant to lay down such a rule. It only stressed the importance of the performance of a witness in the relevant parade, but it issued no directions that the other parades which are held previously or near about the time of the relevant parade should be ignored. The rule of approach to the evidence of identification is clearly stated in : AIR1953All314 . The learned Judges observed :
'Normally the result of identification proceedings in which a particular accused was put up must alone be taken into consideration in deciding this question. It is upon the basis of it that it must be held whether, identification is good and reliable or fit to be discarded. Other identification proceedings in which the particular accused was not put up for identification and other accused were put up are immaterial except in so far as an inference may be drawn against the power of observation of the witnesses. This inference may be drawn from these other identification proceedings when they were held either prior to the identification proceeding relating to the particular accused or simultaneously with or shortly after it. But no conclusion can be drawn from these other identification proceedings if they were held long afterwards, because by the lapse of time a witness may lose that freshness of impression which he might have retained at the rime when the proceeding connected with the particular accused was held. Therefore, identification proceedings held long after the particular proceedings with which the court is concerned should not be taken into consideration in weighing the evidence of identification with regard to a particular accused.'
It is surprising that in spite of the clear language of this decision the trial courts misinterpret and misapply it.
32. Testing the evidence of the two identification witnesses against Babu in the light of their earlier performance we find that P.W. 10 Baldeo failed to identify any one correctly in the first parade held on 9-4-1953 and committed two mistakes. This performance together with the fact that he was not sent to the second parade held on 1-5-1953 leaves little doubt in our minds that the memory of Baldeo was revived by external aid when he went to the third parade and picked out Babu appellant.
33. For all the reasons mentioned above, we set aside the conviction of all the three appellants Anwar, Faqirey and Babu under Section 395, Indian Penal Code and acquit them. They should be released forthwith unless wanted in connection with some other case.