GOSWAMI, J.

1. The day, April 4, 1966, broke ominously for Lala Barhi (deceased) who used to render physical training and swimming lessons to young boys. One such boy, Sabir Hanfi (PW 9), aged about 18 years, went to the house of Lala Barhi (hereinafter, Lala) at Purani Bazar, in the town of Muzaffarpur. Lala was then asleep. Sabir Hanfi woke him up and they both went to the Ashram Ghat (known also as Balu Ghat) on the bank of the Gandak river. There when Lala was cleansing his teeth and washing his face., the appellant Rajendra Prasad (hereinafter to be described as the accused) came there with four or five person s. It is said that the accused had some differences with Lala over some money which he had given to him to assault somebody which Lala failed to accomplish. As his companions were keeping Lala engaged in talk, the accused thrust a dagger on the back of Lala who then called Sabir Hanfi. Lala himself a robust young man. rushed towards the accused who took to his beels with his companions. Lala fell down rushing forward a space. of about forty yards and breathed his last. Sabir Hanfi and others also ran behind Lala to his aid.

2. Although thirteen witnesses were examined by the prosecution only four of them were eye-witnesses to the occurrence. They are Ram Pukar Sah (PW 1), Parmeshwar Prasad (PW 4), Lachman Prasad (PW 10) and Sabir Hanfi (PW 9). The Sessions Judge disbelieved all the eye-witnesses and acquitted the accused. On the other hand the High Court relied on the evidence of PW 9 as being corroborated by PWs 1 and 10. The High Court, therefore, set aside the acquittal and convicted the accused under section 302, Indian Penal Code, and sentenced him to rigorous imprisonment for life. Hence this appeal under section 2(a) of the Supreme Court (Enlargement of Criminal Appellate Jurisdiction) Act, 1970.PW 4 who lodged the first information report without naming any accused and who did not know the accused before the occurrence could not even identify him at the Test Identification Parade which was held on September 29, 1966. The evidence of PW 4 is, therefore, of no significance. The case, therefore, depended upon the evidence of recognition of the accused while running from the place of occurrence by PWs 1, 10 and the direct testimony of PW 9 who knew the accused from before. The accused was known only to Sabir Hanfi (PW 9) by name. The other two witnesses did not know the accused from before and saw him only while running away followed by twenty or twenty five other persons.

3. The evidence of PWs 1 and 10, which we have carefully perused, go to show that they did not know the accused from before. They however, identified the accused in the Test Identification Parade held on September 29, 1966, nearly six months after the occurrence. There is no reason why the Test Parade was delayed so long when the accused had surrendered on June 17, 1966. As stated earlier, the trial court which had opportunity to see these witnesses disbelieved them by giving certain reasons. For example, according to the Sessions Judge, these two witnesses had the opportunity to know the accused from before and, . therefore, their identification in the Test Identification Parade was not considered of such significance. He further observed that these witnesses saw the accused at the time when he was running away from the place of occurrence and, therefore, "it is highly improbable that they would be able to retain the impression of the assailant of Lala Barhi in their mind for so many months". The High Court, on the other hand, held that identification by PWs 1 and 10 was reliable. Thus when two courts have come to a different conclusion on the same evidence, we had ourselves to go through the entire evidence carefully in order to see whether the appreciation of the evidence by the Sessions Judge was so unreasonable anti unrealistic as to entitle the High Court to interfere with the same. PW 1 is the father of PW 10. His uncle died on the previous night and both he and his son (PW 10) went for his cremation at Chandwara Ghat. The cremation was over at about 6.16 A.M. on the morning of the day of occurrence. They then went to Balu Ghat for a bath. They saw Lala Barhi doing physical exercise on the bank of the river and they went to take their bath in the river. After about ten minutes they heard a huIIa `pakro `pakro. On hearing the hulla they came up to the Bank from the river and saw Lala Barhi running away towards the East and four persons were chasing him and raising a huIIa `pakro `pakro. They further stated that about twenty or twenty-five persons followed to catch two or three persons who were chasing the Lala. They had come out of the water at that time. PW 10 stated that he could not identify the other companions of Lala because he "saw their back only and not their face". Since both the witnesses came out of the water on hearing the hulla which was raised only after the assault, it was not possible for these two witnesses to see the stabbing. Even if they recognised the persons running away, they would not be able to say who, amongst them, had stabbed the deceased. PW 10 has candidly admitted that he could not recognise the companions of the deceased as they were running ahead and he was watching from behind, If this be the position with regard to the companions of the deceased it is difficult to see how these two witnesses were able to recognise the accused who was running away. Apart from that there were twenty to twenty five others following the accused with two or three persons thus obstructing a clear view of the accused. In this state of evidence it is impossible to hold, on their evidence, that the learned Sessions Judge did not reac h a correct conclusion that identification by these two witnesses was futile for the purpose of establishing the charge against the accused. We are unable to appreciate how the High Court in this state of evidence could rea sonably come to a contrary conclusion with regard to the testimony of these two witnesses.This leaves PW 9 for consideration. Although PW 9 knew the assailant by name, there is no mention of the accuseds name i n. the first information report lodged the same day at 12.00 noon, five hours after the occurrence, the Thana being only one and a half miles from the place of occurrence. The accused is described in the first information rep ort as "a dark complexioned healthy young man, about the age of 24/25 years". It may be that PW 9 is the only witness who had seen the stabbing. He may be the witness from the beginning of the occurrence till the en d and who described the incident in detail. According to him while the deceased was cleansing his teeth the accused came near him with four or five persons and that while some of then kept the de- ceased eng aged in some conversation the accused stabbed the deceased on his back with a dagger. The whole case will, therefore, rest on the evidence of this witness who knows the accused from before. The Sessions Judge has considered him as unreliable for the following reasons:--

"(1) It is highly improbable that if he had seen the incident he would not have rushed to the police station or even stayed at the place of occurrence till the arrival of the police.

(2) That on the other hand PW 9 confined himself in his house from 4th April till the evening of the following day when a constable went t6 his house to take him to the police station.

(3) The police at first suspected that he might have a hand in the murder of the de- ceased and suspected him and kept him in the police lock-up.

(4) The conduct of the witness is very suspicious and no reliance can be placed on his evidence."

4. The High Court has considered the first two grounds as insufficient for holding the witness as unreliable. It observed that "it is of common knowledge that generally people try to avoid becoming an informant and to be an eye- witness of the occurrence for various reason". With regard to the third and fourth grounds the High Court referred to the station diary entry (Ex. 4) and to the evidence of PW 12 and held as follows:--

"... it was by mistake that PW 9 was arrested by a constable. P.W. 12 has stated that in fact he was a prosecution witness. I do not find any valid reason for discrediting the evidence of P.W. 9".

From the evidence of PW 9 and PW 12 and in the context of the station diary entry (Ex. 4), the position emerges as follows:--

A police officer went to the house of PW 9 in the evening of April 5, 1966, to bring him to the Thana. He w as found by the Officer-incharge of the Thana (PW 12) sitting at the police station at 7.30 P.M.. Now let us read the station diary entry (Ex. 4). Before we quote the same we should state that this exhibit has not been correctly translated at page 56 Of the Paper Book. We, therefore, ourselves examined the original station diary entry and we will set out the same as correctly translated as under:--

"XVIII. Entered in Thana Daily S. 186 186. That this time, Shri G.S. Chaturvedi, Town Inspector, came to Thana and took with him Mohd. Sabir Hanfi alias Funna r/o Saraiyaganj Thana Town---the accused (Abhiyukta) in connection with S. No. 5(4)66, Sec. 302 IPC. On search nothing was found on his person except clothes he was wearing. After all the rules of Hajat were observed he was kept in Hajat".

Some uncertainty was felt by counsel regarding the meaning of the word "hajat". We have no doubt that the word "hajat " means custody in this context. Bhargavas Standard Illustrated Dictionary, Hindi Language, gives the meaning of Hajat, inter alia, as custody, and states thus: "Hajat mein rakhna" means "to keep a culprit in custody".

5. The High Court, we are afraid, does not appear to have examined the original station diary entry (Ex. 4) with care otherwise it would not have come to the conclusion that it was by mistake that PW 9 was arrested by a constable simply, from the self-serving statement of PW 12. The High Court does not even refer to the fact that a very important recital in the original entry (Ex. 4), namely, the word "abhivukta" (accused) ha s been scored out and in its place the word "gavah" (witness) was substituted. The interpolation of the word "gavah" (witness) for "abhiyukta" (accused) appeared to be of a different writer from the original writings in the entry. This is serious tampering with an official record in a criminal case when a mans life is in jeopardy in a trial for murder under section 302 IPC. We have also grave suspicion about the date of correcti on of the entry in Ex. 4.Although the above entry shows that PW 9 was brought to the Thana by the Town Inspector, G.S. Chaturvedi, he was not even examined in the case to show why and under what circumstances PW 9 was brought by him. This entry in the official record clearly shows that PW 9 was at that stage considered as an accused in connection with this murder case and his person was searched before confining him in the Hajat. If he was only a witness there was no reason why his person would be searched. and why he would be kept in the lockup "after all the rules of Hajat were observed". The High Court completely lost sight of these serious infirmities in the prosecution evidence and it was absolutely impermissible to accept the statement of PW l 2 when he stated that PW 9 "had not been arrested" and that "he was only a prosecution witness.

6. There are some other disquieting features with regard to the evidence of PW 9. It is not understood why he should have said that a constable brought him to the Thana whereas the entry (Ex. 4) shows that the Town Inspector, Chaturvedi, brought him to the Thana. If, as the entry shows, he was brought by the Town Inspector, Chaturvedi, and he was kept confined in the lock-up as an accused in the murder case, it is difficult to comprehend how at that very stage Ex. 4 could be considered by PW 12 as containing a wrong recital without reference to the Town Inspector. Besides, PW 12, as he says, had taken charge of the investigation from Sub Inspector, Gupteshwar Dayal (PW 13) at 11.00 A.M. on April 5, 1966, inspected the place of occurrence, unsuccessfully searched for the suspects Rajendra Prasad, Ram Bilas Sahani and Mohan Jha at their houses examined some witnesses and then reported to the Thana at 7.30 P.M. to find PW 9, the principal witness, sitting there. Even then he would not record the statement of PW 9. He admitted during the course of cross-.examination that when he first interrogated PW 9 at the Thana he had not recorded his statement. He did not even record his statement when he came to the Thana from his house at 11.00 P.M. that night, These are very suspicious circumstances. PW 12, however, admitted that he recorded the statement of PW 9 on the following day (6-4-1966) at 12.05 A.N., after having pro- duced him before the Superintendent of Police. PW 9 admits that he was interrogated by the Superintendent of Po1ice for twenty to twenty five minutes. PW 12 stated in his examination-in-chief that he allowed PW 9 to go to his house after h e had recorded his statement on April 6, 1966, at 12.05 A.N. From the evidence of PW 12 read with the entry (Ex. 4) it does not appear that PW 9 was freed from police custody at least till his statement was recorded on April 6, 1966. Again, the statement of PW 9 was recorded by the Magistrate under section 164, Criminal Procedure Code, on April 12, 1966, when perhaps the police had finally decided to treat him as a prosecution wit ness instead. of an accused. This conclusion is irresistible on the state of evidence to which we have referred above. If under these circumstances the Sessions Judge held that the conduct of this witness was such as would seriously affect his credi- bility, the High Court was not at all justified in taking a contrary view even without a proper analysis of the oral and documentary evidence. When the evidence of recognition of the accused by PWs 1, 10 and 4 is unreliable, no conviction can be based on the sole testimony of a witness like PW 9, on whom the first suspicion fell, without any corroboration..When a trial court, with full view of the witnesses, acquits an accused after disbelieving direct testimony, it will be essential for the High court, in an appeal against acquittal, to clearly indicate firm and weighty grounds, from the record, for discarding the reasons of the trial court in order to be able to reach a contrary conclusion of guilt of the accused. The High Court should be able to point out in its judgment that the trial courts reasons are palpably and unerringly shaky and its own reasons are demonstrably cogent. As a salutary rule of appreciation of evidence, in an appeal against acquittal, it is not legally sufficient that it is just possible for the High Court to take a contrary view about the credibility of witnesses but it is absolutely imperative that the High Court convincingly finds it well-nigh impossible for the trial court to reject their testimony. This is the quintessence of the jurisprudential aspect of criminal justice.

7. This is not a case where it can be said that the Sessions Judge came to a palpably wrong conclusion on the evidence or that the reasons for acquittal of the accused are manifestly erroneous, shocking ones sense of justice. The High Court was not right in interfering with the acquittal of the accused in this case. The appeal is, therefore, allowed. The judgment of the High Court is set aside. The accused shall be released from detention immediately.

8. Appeal allowed.