1. The appellants in this case are 31 in number. They have been tried in the Sessions Judge's Court of Benares for offences under Sections 148, 302/149 and 307/149 and 147, Penal Code. Gokul, Nazir Pancham, Sumer, Shukul and Tika have been convicted of rioting armed with a deadly weapon, of murder, and of attempted murder. Bechu, Babban, Baijnath, Bahoran, Bageshri, Buchnu, Damri, Gharbari, Harinandan, Hira Lal, Jeot Koeri, Kallu Chamar, Katholan, Kalicharan, Misri Lal, Mahesh, Morahu, Mauj Brahman, Panaru Kohar, Ram Kishen, Ramman, Sadho, Sheo Prasad, Saggam and Shukalu have been convicted of rioting, murder and attempted murder. Gokul, Pancham, Tika, Sumer, Shukul, Marahu and Saggam have been sentenced to death under Section 302, Penal Code, and to transportation for life under Section 307, Penal Code; Gokul, Sumer, Pancham, Shukul and Tika to three years' rigorous imprisonment and Marahu and Saggam to two years' rigorous imprisonment under Section 148, Penal Code. Nazir and Kutballi have been sentenced to three years' rigorous imprisonment each under Section 148, Penal Code, and to transportation for life under Sections 302 and 307, Penal Code. Bechu, Babban, Baijnath, Bahoran, Bageshri, Bachnu, Damri, Grmrbari, Harinandan, Hira Lal, Jeot Koeri, Kallu Chamar, Katholan, Kali Charan, Misri Lal, Mahesh, Mauj Brahman, Panaru Kahar, Ram Kishen, Ramman, Sadho, Sheo Prasad and Shukalu have been sentenced to transportation for life under Sections 302 and 307, Penal Code, and to two years' rigorous imprisonment each under Section 147, Penal Code. The offences with which the accused have been charged and convicted are alleged to have been committed in the village of Raiya, P.S. Punnuganj on 16th March 1931.
2. The riot in which four persons were killed and two seriously injured was the result of bitter communal antagonism in the district. For some considerable time before the riot hostility between the Mahomedans and the Hindus had been smouldering. Two days before, that is, on 14thl March 1931, this hostility flared up in the neighbouring village of Manchi where a riot occurred, the Hindus attacking the Mahomedans. It is unnecessary to go in detail into the events leading up to this not. These are described by the learned Sessions Judge of Benares in his judgment in the Manchi riot case. Suffice it to refer to the fact that feeling for some time between the two sections of the community had been running high and that the immediate cause of the outburst was the alleged killing of a cow by one Mohammad Raza. The Hindus were determined to avenge the sacrilege and the riot in Raiya in which the accused are alleged to have participated was really a continuation of the riot in Manchi on 14th March 1931. It appears that the Hindus had determined to attack the Mahomedans of village Raipur which is about one mile distant from the village of Raiya because the leading Mahomedan in Raipur, Imam Bakhsh, gave shelter to Mohammad Raza who was said to have killed the cow and who succeeded in making good his escape from Manchi on the 14th.
3. Imam Bakhsh received a warning from the witness Tengar of the impending attack by the Hindus of Raipur on the 15th March. It is stated that this witness acted as an intermediary between the Mahomedans and the Hindus who demanded from Imam Bakhsh payment of Rs. 400 as the price of ithe safety of the Mahomedans in Raipur. If this sum were not paid, he informed Imam Bakhsh, the Hindus were determined to plunder and kill the Mahomedans in Raipur. On the morning of the 16th March the witness, constable Dip Narain Singh, who was aware of the impending attack by the Hindus arrived in Raipur and advised the Mahomedans to leave the village. At about 10 a. m. the Hindus were seen advancing in force from the village of Khalyari which is about 2 miles distant. On their approach the Mahomedans fled from Raipur taking with them their women and children. They were pursued by the Hindu mob and upon reaching Raiya Imam Bakhsh directed them to take shelter in the house of Raghunandan Brahman who was on friendly terms with the Mahomedans. The refugees with the exception of the witness Shukurullah Saham Ali and Ramzan Ali, forced their way into Raghunandan's house. Shukullah, Saham Ali and Ramzan endeavoured to make good their escape from the village. Saham Ali and Ramzan however were overtaken and killed and later in the day their bodies were burnt; only their charred bones were eventually recovered. Shukrullah succeeded in making good his escape. He made his way to the police station at Punnuganj where he made the first information report at about 4 o'clock in the afternoon.
4. It is unnecessary here to recapitulate the story of the attack on the house of Raghunandan by a Hindu mob. That has been related in detail in the order of the committing Magistrate and. in the judgment of the learned Sessions Judge. It is sufficient to say that the refugees were eventually rescued by armed police who arrived on the scene. Before the police came however Lal Mohammad and Gaffur had been killed by gun-shot, Farzand wounded by gunshot and Majid beaten and rendered unconscious. A large number of the Hindus were rounded up by the police in the village of Raiya and were takeri to the thana. The list of their names is Ex. Y. From these persons the police collected a large number of weapons: axes, spears and lathis. No attempt has been made however to prove that any of these weapons was the property of the accused. A hundred Hindus were subsequently challaned by the police. Ninety-nine appeared before the committing Magistrate who discharged 16 and committed 83. During the course of the trial before the Sessions Judge one accused died. The Sessions Judge acquitted 50 of the charges which were preferred against them and convicted 32. Seven of these have been sentenced to death and the remainder to transportation for life.
5. This case presents certain special features to which in view of our decision we desire at this stage to make reference: (1) In this case the prosecution has relied almost entirely upon the oral testimony of the prosecution witnesses. Their evidence has not been supplemented or supported by facts and circumstances such as, for example, the recovery of arms or blood-stained garments. As has been stated above no attempt was made to identify any of the weapons taken possession of by the police as the property of the accused and although firearms are alleged to have been freely used, none appear to have been recovered. (2) The Pandeys of the village of Khalyari who have been convicted and sentenced are all related, being descended from a common ancestor. None of the Pandeys of Khalyari, who, according to the witnesses, took a prominent part in the riot were rounded up by the police in Raiya and taken to the thana and their names do not therefore appear in Ex. Y already referred to. This is an extraordinary circumstance. (3) The prosecution witnesses are practically all Mahomeedans and many of them are related to Imam Bakhsh, the leading Mahomedan in Raipur where he is known as the Bara Babu. We append to our judgment a chart showing the relationship of these witnesses. (4) Between many of these witnesses and a number of accused there had been long-standing enmity, an enmity which is evidenced by numerous civil and criminal actions. We append to our judgment a chart setting forth these actions. If litigation is proof of enmity between witness and accused, enmity, bitter and long standing, has been established in this case. (5) Two reports were made to the police purporting to give the names of those who took part in the riot. One report is by the witness Ismail, son of Imam Bakhsh, (Ex. 12) and the other by the witness Ali Husain, brother of Imam Bakhsh (Ex. 6). The persons named in these reports, Ismail and Ali Husain, say they recognized during the course of the riot. Both reports include a very large number of names and it is impossible to believe that the witnesses recognized and remembered so many of the rioters. Ismail names sixty persons, Ali Husain ninety. Both reports were made to the police two days after the riot.
6. The facts above referred to are sufficient to rouse the suspicion that the Mahomedan witnesses may have falsely named a large number of innocent persons. During the course of the hearing of these appeals this suspicion was confirmed by the appearance in Court of the appellant Shukul who had applied to be present. (After examining and discussing the evidence as against Shukul, the judgment proceeded). We now consider the bearing of the clearly false implication of Shukul upon our decision in the appeals of the other accused. So far as Shukul is concerned we unhesitatingly acquit him finding that the case against him is false. Seventeen witnesses have sworn on oath that he took an active part in the riot. These 17 witnesses are guilty of perjury. They have given false evidence against one accused charged with murder. We have come to the conclusion that we cannot rely upon their evidence as against the other accused. The 17 witnesses who have testified against Shukul are Imam Bakhsh, Ra-ghunath, Shukrullah, Ali Husain, Taj Uddin, Mt. Habiban, Mt. Rajmati, Raz-zaq, Tafazzul, Abhaiman, Mt. Idan, Sita Ram, Gaya Koeri, Ramdhani, Sukal, Jhuri and Farzand. We refuse to believe their testimony against any of the accused. For the same reason, we refuse to accept the testimony of Mt. Bachchi, who gave evidence against Shukul in the Manchi riot case and whom the learned Sessions Judge in this case regards as a good and reliable witness. Mt. Bachchi has clearly been guilty of perjury against Shukul in the earlier case. We decline to accept her testimony against any of the accused in the present case. (After discussing the value of the prosecution witnesses, the judgment proceeded). For the sake of convenience we append to our judgment a chart prepared at our request showing the criticisms of the learned Sessions Judge of the witnesses in this case. On the evidence of these witnesses the learned Sessions Judge has acquitted in some cases but convicted in others. The result is that we are compelled to allow all the appeals and set aside all the convictions and sentences. No doubt some guilty persons are thus set free and some murders go unpunished, but to find any particular person guilty in this case upon such evidence would be a sheer gamble.
7. In view of the circumstances of this case and of the decision at which we have arrived we consider it expedient to make certain general observations with regard to the administration of justice: (1) The duty of the police in the investigation of any crime is to discover the truth and not simply to obtain evidence for the purpose of securing a conviction.
(2) It is the duty of the prosecution to bring out in evidence everything in favour of an accused person and to lay before the Court all the evidence even though some of that evidence may result in an acquittal.
(3) It is the duty of the Committing Magistrate and the trial Judge to be solicitous in the interests of the accused. This is especially so in riot cases where the accused are generally humble and ignorant people unable to defend themselves, and often inadequately represented in the Courts. (Under Indian law they are not even allowed to give evidence for themselves. If Shukul could have been put in the witness-box the whole aspect of the case for the prosecution would have been changed). Had the committing Magistrate and the learned Sessions Judge in this case observed this rule the accused Shukul Pandey would never have been committed for trial or convicted. We recognize the difficulties of the Judges in the lower Courts especially in a case such as the present where there is a large number of accused. But it is just in these cases that Judges should take great pains to examine the evidence minutely in considering the prosecution case, with a view to protecting the accused. We would state further that it is emphatically the duty of the Magistrate who commits or the Judge who convicts to see each of the accused before he commits or convicts and passes sentence. If the learned Sessions Judge had appreciated the condition of Shukul Pandey we are convinced he would never have convicted him and sentenced him to death. It is to be noted that in the lower Courts counsel for Shukul never drew the attention of the Court to the condition of his client. The point was taken for the first time in this Court. This is one of the amazing facts of this case.
(4) We are of opinion that in many cases sufficient importance is not attached to the identification of accused by witnesses in the trial Court. In the present case it appears that in certain instances a witness was unable to pick out a particular accused he had named from amongst the accused. When this occurred, the accused person or persons whom it was desired that the witness should identify were asked to stand up, and in one case, at any rate, made to give their names. The witness then identified the accused.
8. This procedure reduces identification to an utter farce. Identification by a witness thus obtained is worthless. We note here some examples of the method that has been adopted in the Sessions Court:
(i) (Picks out all except Dangai and Saggam's brother, Hira Lal, and Saggam. Also at first he picked out Mahesh for Sadhu but then picked Sadhu correctly). (Dangai, Saggam and Hira Lal made to stand up). These are Dangai, Saggam and Saggam's brother. (Picks them out correctly)). (Vol. 2, p. 144, lines 41 to 44).
(ii) (Picks out Morahu only). Before the lower Court I recognized only Morahu and Baijnath. Now I only recognize Moharu. (Baijnath made to stand up). This is Baijnath. (Vol. 2, p. 102, lines 1 to 3).
(iii) (Witness fails to identify accused). (Panaru, Misri and Banwari made to stand up and give their names). Q. Do yon recognize these men? A. 'Yes'. (Vol. 2, p. 79, lines 44 to 47).
9. We are amazed that such procedure was tolerated by the Sessions Judge or allowed to pass without obejction or comment by counsel for the accused.
5. It was argued by counsel for the Crown that even though the 'Shukul' witnesses had committed perjury in their evidence against Shukul their evidence ought to be taken into consideration against other accused. It was contended that this was in accordance with the general practice to the Courts in India: that in most criminal cases false evidence was given on one side or the other or even on both: that if prosecution witnesses were discredited as against all the accused because they had committed perjury in the case of some accused prosecutions in India would seldom be successful and criminals would escape punishment.
10. It is unfortunately true that in India it is often the practice for complainants to implicate their enemies falsely in criminal cases. This, of course, makes the trial of accused persons a matter of great difficulty and anxiety to all Judges. We think however that the argument of counsel is too broadly stated. It is seldom that the Court can come to a definite conclusion that perjury has been proved to have been committed by particular witnesses. The correct procedure in our opinion is that in cases where there is reason to believe that certain accused, on the ground of enmity or otherwise, may have been falsely charged then the evidence of those witnesses who have reasons falsely to implicate, the particular accused should not be relied on as against that particular accused: on the other hand, the same witnesses might be relied upon against other accused where there is no reason to suspect enmity on the part of the witnesses. Where however perjury has definitely been brought home to a witness it would be extremely dangerous to rely on his evidence against anyone. This proposition is frequently stated as follows:
We do not think there is much difference in this way of stating the proposition. The effect is the samp: the evidence of the witness is in reality not relied upon. We cannot see why such a witness should be believed on oath at all. This case brings into sharp relief the futility of the oath as at present administered to Indian witnesses. We have frequently been assured by Indian counsel of great experience that this eath is in the case of many Indian witnesses of no binding effect; and indeed that is our own experience. There is a common phrase in this Province which clearly illustrates this. 'Suck bolo adalat men nahin ho,' that is 'Speak the truth. You are not now in a Court of law.' There is behind the present oath no religious sanction Further, witnesses know that Courts hesitate to put the law in motion against perjurers. The Courts in India are so greatly in arrears with work as it is and the increase of work, if prosecutions for perjury were regularly instituted, would be extensive. There is thus little penal sanction behind the oath either. A binding oath with religious sanction behind it: is one of the foundations of the administration of justice. We think if there were such an oath in India-especially in the case of villagers who provide 90% of the witnesses in criminal cases-perjury would be no more prevalent than elsewhere; and the difficulty in trying such cases would largely disappear. The pressure on the Courts of Justice of the mass of litigation of all lands would also be reduced. Fewer false cases would be brought, and, if brought, fewer witnesses would be found to support them. We now deal with the cases of individual appellants.
11. In considering these appeals for the sake of brevity we shall refer to the witnesses who give evidence against Shukul Pandey as 'Shukul' witnesses. (After dealing with them individually, the judgment concluded.) In the result, for the reasons given, we set aside the convictions and sentences of all the appellants and direct that they be set at liberty forthwith.