1. Seven persens, Nazir, Jumma, Shera, Idu, Sukhan, Harbans and Pheru, have been convicted by the Sessions Judge of Muzaffarnagar of an offence under Section 396, I.P.C, and have been sentenced each to 8 years' R.I. The prisoners have submitted appeals from jail against their conviction, and an application has been made in revision on behalf of the Local Government requesting this Court to enhance the sentences passed upon these persons.
2. A dacoity of a ferocious nature took place on the night of 19th/20th October, 1930, at the house of Prabhu Lal Vaish in the village of Kaserwa.the dacoits, who are said to have numbered more than 30, were armed with guns, lathis and probably spears. Two men who interfered with them were shot dead, and another was beaten to death with lathis. The women were grossly ill-used, one or more were violated, and one was tortured. In my opinion a sentence of 8 years' R.I. in such a case is inadequate. No sentence less than transportation for life should be imposed upon any, of those persons who take part in a dacoity of this nature. It does not appear that the learned Sessions Judge had any doubt that the persons whom he convicted were guilty, and he has praised the police officers who conducted the investigation for their good work. We regret that we have been unable to agree with the learned Judge either in his commendation of the police investigation or in the view taken by him as to the guilt of the majority of those persons whom he has convicted.
3. The case was reported by Faqira chaukidar at 8 a. m. on 20th October. The 1st portion of the report is a straightforward description of the affair. At the end a suggestion is made that the dacoity was the work of the congress party organised by the congress captain Lal Singh jar. This part of the report does not appear to be the work of the chaukidar, and we are not surprised to find that the chaukidar denied emphatically that he had made this addition.
4. It appears that at that time there was much political trouble in this police circle, and it may very well be that the police officers who heard the report made thought that the dacoity might be the work of the congress party, and they clearly commenced the investigation on this supposition. It appears that they were mistaken, and it is only fair to the police officers concerned to say that they did not persist in attempting to implicate members of the congress party when they found that there was no evidence in support of this view. (The judgment then narrated the progress of investigation upto the identification of the accused and proceeded.) On 3rd February identification proceedings were conducted in jail by a Magistrate of the first class, Mr. Gobind Prasad Mathur. The persons pufi up for identification were Nazir Jumma, Nahar, Risal, Fateh Muhammad, Shera, Idu, Sukhan, Jhandu, Harbans, Pheru, Bias, Baru and Aiman. the result of the identification is given in the report made by the Magistrate at the time. It is true that all the present appellants were identified by one or other of the witnesses, and the learned Sessions Judge has accepted these identifications as being evidence against them.
5. There is nothing in the judgment to show that the identification is rendered valueless by the fact that the witnesses in every case picked out more persons who were not suspected than persons who were suspected, and in most cases were so reckless in their identifications that they might well have made them blindfold. Thus Prabhu, the complainant, identified two of the accused persons and 14 who were not connected with the affair; Dalip identified 5 accused persons and 21 others; Fateh Singh identified 8 accused persons and 14 others. Those who were less reckless identified persons who have been acquitted, notably Fateh Muhammad, who is stated in the first confession of Ilias to have remained in a grove and not to have gone to the village, and has been acquitted by the Judge on the ground that he was implicated falsely out of enmity; and the best of all the witnesses Faqira made 4 mistakes as against 3 identifications of persons who were suspected. The fact is that the identifications were carried out in circumstances that made a trustworthy identification very difficult. There were 15 persons put up for identification and 80 others. It is better in such cases to divide the numbers and have separated identifications of 10 or 15 persons containing two or three of the accused.
6. In my opinion these identification proceedings are almost, if not quite, negligible. Faqira, the chaukidar, is the only one of the witnesses on whom I could place even the slightest reliance. Another witness Hardeo appears to have identified 5 persons suspected and 8 others. But it appears certain that of the 5 he had identified he knew Harbans before, and Fateh Muhammad was probably not present at the dacoity. Consequently his identification is of little value. But it is to be noticed that both he and Faqira identified Idu and Jam-ma, and that fact may be taken as very slight corroborative evidence against those persons if there is sufficient evidence of a different nature against them. The following day the same Magistrate held an identification of property recovered from the accused persons. (The judgment considered the evidence on this point and proceeded.) Md. Ilias was offered a pardon and examined on on oath as an approver. The Judge is of opinion that Ilias has falsely implicated his personal enemy Fateh Muhammad, and it is easy to show from his statement that all the details that he has given as to the actual dacoity have been added to the original statement which he made before the Magistrate. The Judge has mentioned certain particulars from which he considers that the approver's statement has been corroborated. The first corroboration is found in the fact that he showed to Mr. Ganga Ram Yadav, Deputy Magistrate, the different places where the dacoits had collected and where he was standing at the time the dacoity was being committed. As no one saw Ilias at these places it is difficult to see how his statement has been corroborated by the fact that he showed certain places to the Magistrate. Secondly, the learned Judge says that his account of this dacoity is corroborated by other dacoity witnesses. This is true of his account as given in the Sessions trial, but it is not true of the statement made by him before the Magistrate on the 15th December, 1931, in which he gave no details. Thirdly the Judge states correctly that his statement is corroborated by his production of a Jhanwar and six chharas which, as I have already shown, correspond to articles mentioned in the list given by the complainant on the 21st October. Fourthly the Judge refers to a confession made by Nazir on the same day, to which I shall refer later; and lastly, he has referred to the identification of Ilias by certain witnesses.
7. On this identification I have already commented. In my opinion the statement of the approver has very little corroboration. There is enough to show that the man took part in the dacoity himself, but as a witness against others, the view taken by the Judge that he has, falsely accused Fateh Muhammad and. Jhandu must be considered as weakening the general effect of the statement, A reference to the statement will show that this man has in Court added definite accusations against other members of the gang which were not made by him in the first instance. In particular there is his statement that Jhandu and Harbans fired the fatal shots at the two persons who were killed, and there is abundant evidence that the approver sought to improve his statement by adding details in Court which he did not give at the time of his first confession.
8. The statement, however, is evidence admissible against the persons whom he has named subject to corroboration by other evidence. Of the appellants before us Ilias in his first confession named only Nazir and he mentioned a Jat with a bicycle and a gun. The other persons who appeal were neither mentioned nor indicated by Ilias in his first confession. (The judgment then considered the cases of Jumma-Sukhan, Pheru and Shera, and after holding that there was no sufficient corroboration against them, proceeded.) There remains the case of two persons who made confessions and retracted them. I have already stated the circumstances in which Nazir made his confession. It appears that it was a confession made on an inducement by some person in authority and as such is inadmissible in evidence under Section 24, Evidence Act. Section 28 cannot be applied in this case because there is every reason to believe that the warning conveyed by the Magistrate did not remove the impression caused by the inducement. If his own confession is not admitted in evidence against him there remains only the statement of Ilias and the confession of Idu. In my opinion no reliance can be placed on Idu in respect of the names given by him. Consequently when he says that Nazir took part in the dacoity this is not a statement which can be regarded as evidence against Nazir. The ornament found in this man's possession is not proved to have been connected with the dacoity, and the uncorroborated evidence of the approver is not sufficient to justify a conviction. The only person who identified Nazir was the witness Qabul dhobi who picked out in jail 4 suspected and 18 unsuspected pesons. Seeing that Nazir is a retainer of Ilias and was clearly acting throughout under his influence and suggestion I cannot say that I have no doubt as to his guilt in spite of his confession and I consider that he is entitled to an acquittal.
9. There remains only Idu. I cannot find any reason for discarding Idu's confession as against himself. It was made voluntarily and at his own express wish because he hoped to be made an approver, but I am satisfied that no promise was ever made to him that he would be an approver, and he gives details in his confession which clearly indicate that he was one of the decoits. Moreover the two least unreliable witnesses both identified him in jail. In my opinion therefore all of the appellants except Idu must be acquitted. In the case of Idu I would accept the Govt.'s application for revision and enhance his sentence to transportation for life. It is unfortunate that in a case of this kind only one man should have been brought to justice, but this is bound to happen so long as the investigating officers are satisfied when they obtain a confession and base their case on the confession irrespective of the fact that the man who has made the confession is in no way a reliable person and has done nothing to assist in the proper investigation of the case. In my opinion Ilias should never have been offered a pardon after the investigating authorities had time to consider the effect of the statement made by him before a Magistrate and the circumstances in which it was made. I do not wish to criticise too severely the learned Sessions Judge who has no doubt tried to come to a proper finding in the case and has honestly believed that the persons whom he has convicted were guilty, but it is the duty of the Judge in all cases to state fairly the evidence both for and against each of the accused persons, and this the Judge has not done. However anxious a Judge may be that his decision should be upheld on appeal he should never attempt to secure his aim by making the case against an accused person appear to be stronger than the evidence justifies.
10. That one Pirbhu Bania of village Kasarva Kalan in the district of Muzaffarnagar was a victim of a serious decoity committed in his house on the night between 19th and 20th October 1930 is established beyond doubt. Some of the dacoits had firearms which they used in terrifying the villagers who attempted to come to the rescue of Pirbhu. They actually killed three persens in course of the dacoity. The learned Sessions Judge has convicted seven out of eleven persons tried by him, sentencing each to eight years' rigorous imprisonment under Section 396, I.P.C. I may say at once that the sentences passed by the learned Sessions Judge are wholly inadequate if the guilt of the appellants is established. At the same time, the case is one in which evidence should be sifted with care to determine whether a serious offence of the nature described above has been brought home to every one of them.
11. The principal evidence on which the conviction of most of the appellants rests is the statement of the approver, Ilyas. It is now settled law that unless an approver's testimony is corroborated not only as regards corpus delicti but also as regards the identity of such accused person, conviction is not justified. The finding of the learned Sessions Judge that the approver Ilyas took part in the dacoity is based on sufficient evidence and may be safely accepted. The more important question however is whether all those whom the approver implicates have been proved to have taken part in the dacoity. He must be corroborated as regards each accused, even if his evidence is not open to serious challenge in any material particular, which is by no means true in case of Ilyas. His evidence is somewhat discounted by the fact that, in the confession made by him for the first time on 15th December 1930, he did not name any of the appellants, except Nazir, who also made a confession on the same day but subsequently retracted it. The approver prominently mentioned Fateh Muhammad, who according to him took a ;very active part in organising the dacoity. Fateh Muhammad has been acquitted by the learned Sessions Judge on definitely finding that his name was falsely mentioned by the approver Ilyas out of enmity. In his subsequent statement, made as a witness, he mentioned by name no less than 18 persons including the appellants who according to him were among the dacoits.
12. Besides the evidence of the approver, we have the retracted confessions of Nazir and Idu implicating not only themselves but also their co-accused. For reasons which will appear presently, I agree with my learned brother in holding that Nazir's confession should be excluded altogether not only so far as other accused are concerned but as against himself also. Idu's retracted confession, if found to be voluntary and admissible in evidence, is an important piece of evidence against himself ; but as against his coaccused, the Court can only "take it into consideration." The retracted confession oi a coaccused, like the evidence of the approver, is not sufficient for conviction, unless it is corroborated in material particulars: See Emperor v. Kalwa AIR 1926 All 377. Another question which requires consideration in the present case is whether, if there is no other reliable evidence in corroboration, the evidence of the approver, coupled with the retracted confession of a co-accused, is sufficient for conviction. There is ample authority in support of the view that it is not. It has been expressed in the dictum that one tainted testimony cannot corroborate another tainted testimony. Numerous cases bearing on this point have been quoted in the Law of Evidence by Woodroffe and Ameer Ali under Section 30, at p. 315, 9th Edn., wherein it is stated that.
the confasion of co-prisoners to be rendered trustworthy must be corroborated aliunde by independent evidence and not by the testimony of accomplices or approvers as well in respect of the corpus delicti.
13. It stands to reason and common sense that if the approver's evidence itself is not sufficient for want of corroboration, the retracted confession of a co-accused, which again has been declared to be equally, if not more untrustworthy, does not furnish the requisite corroboration. The two together cannot obviate the necessity of corroboration from independent sources. The corroboration need not be full proof of guilt but should establish a sufficient degree of complicity. The whole principle of the rule which insists on corroboration of the approver or of the retracted confession of a co-accused is based on the assumption that, in the absence of a guarantee forthcoming from more trustworthy sources, the evidence of the approver or the retracted confession or both together cannot suffice for conviction. With the above general remarks 1 shall consider the case of individual accused persons. To take the case of Nazir first: he is said to have been arrested on 10th December 1930. Ilyas was also arrested on the same date. An application was made by the Circle Inspector requesting that their statements be recorded under Section 164, Criminal P.C. Chaudhri Ganga Ram Jadava, Magistrate, accordingly sent for them on the 15th from the jail lock up where they had been kept under the orders of a Magistrate and their statements were recorded. Nazir made a confession much of the same effect as Ilyas did. Ilyas, to whom a pardon was subsequently offered, adhered to his confession. Nazir however retracted it when examined as an accused before the committing Magistrate. In explaining his confession before the Sessions Judge he alleged that he had been subjected to illtreatment by the Sub-Inspector Aslam Khan and that he made the statement which Aslam Khan desired him to make. He examined a witness, Hamid Ali Khan, who deposed that Ilyas and Nazir were taken before the Magistrate for record of their confessions but they refused to make any and that subsequently the Sub-Inspector Aslam Khan induced them to confess, promising to secure a pardon for each of them. Hamid Ali Khan has been disbelieved by the learned Sessions Judge for reasons given in his judgment. 1 am prepared to accept his view in that respect.
14. But I am somewhat suprised at the remark of the learned Judge that "there is absolutely nothing suspicious about this confession." Nazir too, like Ilyas, remained in jail for four days before his confession, and it appears, apart from what the defence witness Hamid Ali Khan stated, that both of them, when placed before the Magistrate for making confessions refused to make any incriminating statement. This was elicited from Chaudhri Ganga Ram Yadava, Magistrate, who was examined as a witness both before the committing Magistrate and the Session Judge. It is to be regreted that Mr. Yadaya did not record any proceedings of what transpired when Ilyas and Nazir were first placed before him for their confessions being recorded. Section 164, Criminal P.C. provides for "confessions" or "statements" of persons willing to make them being recorded. It is clear that the police expected that Ilyas and Nazir would make confessions, otherwise the Magistrate would not have been moved to send for them and record their statements. The Magistrate having agreed to take proceedings under Section 164 and sent for them, it is highly probable that he put the usual questions warning them of the consequences of their making confessions, whereon they refused to make any incriminating statement. It is clearly not the function of the Magistrate, when he is to record statements under Section 164, to reduce to writing what the persons placed before him say only if they make statements which the police expect them to make and otherwise to make no note of anything which they say. The learned Sessions Judge has given some reasons for disbelieving the witness Hamid Ali, who stated that Ilyas and Nazir first refused to make confessions but were subsequently induced to make them; but he makes no mention of the facts stated by Chaudhri Ganga Ram Yadava. I do not think that the Sessions Judge, who himself recorded his statement, could have forgotten what he said, nor do I believe that he failed to appreciate the importance of facts stated by him. It was clearly the duty of Chaudhri Ganga Ram Yadava to have kept a record of what the accused were asked and what they stated on the first occasion when they were placed before him and when they refused to make any confession. As it is, we can only have such details of what happened on that occasion as Mr. Yadava could depose from memory on 26th October 1931, over 10 months afterwards. This is what was elicited from him in cross-examination:
Ilias was sent to jail under a warrant dated 11th December 1980, and on it there is my order dated 11th December 1930, that inasmuch as the statement of Ilias was to be recorded on the next day and so he was to be kept separate from other prisoners. This warrant does not bear my signatures. It bears the signature of K.B. Md. Mustafa Sahib, Deputy Magistrate. I cannot say if the application by the police to record the statement of Ilias was oral or in writing. On 15th December when I wanted to record the statements of Ilias and Nazir, they did not at first want to make any statement. So far as I remember, the police was present outside the Court room. When they did not want to make any statement, they were sent out of, the Court room. A little after, within 5 or 7 minutes, they came back again, of course under the police guard, and they said that they would give their statements. I cannot say what policemen were present outside the Court room. But I remember that on that day, before their statements were recorded, Aslam Khan, S.O. of Shamali, came to meand, asked me if the statements of Ilias and Nazir wore recorded, and I replied that they were to be recorded on that day. During the 5 or 7 minutes that Ilias and Nizar went out of the Court room, they were in the custody of the police guard. I must state here that when they returned to the Court room on the second time, they were given full time to wait and think, and thus only their statements were recorded when they insisted that their statements should be recorded. I did not certainly put the question to them as to why they were making a free and voluntary statement. There is no note of mine that they refused to make any statement and that it was written when they persisted. When they first came and refused to make any statement, they were told to go away.
15. The above statement was recorded by the learned Sessions Judge himself, and if he had applied his mind to all the circumstances, beginning with the arrest of Ilyas and Nazir on the 10th the request of the police that their confessions be recorded, their being kept separate from other undertrial prisoners under the orders of a Magistrate, their refusal to make a confession when first questioned and their readiness after a few minutes, which they spent in police custody, to make confessions and the confession itself in the case of Nazir followed by a retraction when pardon was given to Ilyas and not to Nazir, the learned Sessions Judge would not have made the remark which I have quoted above, namely, that there was nothing suspicious about the confession. To my mind, the inference is irresistible that a promise of pardon had been held out to Nazir and Ilyas, and telying on that promise they agreed to make confessions. The police requested on the faith of their readiness to confess that their statements be recorded. The effect of promise lasted on their minds till they were placed before the Magistrate, whose warning opened their eyes and they refused to make any confession. Subsequently they were persuaded to believe that the warning given by the Magistrate was of a stereotyped and formal character and should carry no significance. Nazir made a confession, but subsequently, when he discovered that Ilyas was the favoured individual and that no. pardon would be given to, him, he retracted. Section 24, Evidence Act, does not require the same cogency of evidence as is necessary to establish a fact. It merely requires that if it "appears" to the Court that a confession was induced by threat or promise proceeding from a person in authority in relation to the charge against the accused, it shall be inadmissible in evidence. It has been held in numerous cases that if circumstances create a probability in the mind of the Court that the confession was improperly obtained, it should be excluded from evidence. In the present case, as already stated, the circumstantial evidence is so strong as to eatablish the fact that Nazir was induced to make a confession by a promise of pardon held out by the police. A very eminent Judge of this Court expressed himself on the subject of retracted confessions in the following terms:
To repeat a phrase I used on a former occasion, instead of working up to the confession they (the police) work down from it, with the result that we frequently find ourselves compelled to reverse convictions simply because, beyond the confession, there is no tangible evidence of guilt. Moreover I have said, and I repeat now, it is incredible that the extraordinary large number of confessions which come before us in the criminal cases disposed of by this Court, either in appeal or revision, should have been voluntarily and freely made in every instance as represented. I may claim some knowledge of, and acquaintance with, the ways and conduct of persons accused of crime, and I do not believe that the ordinary inclination of their minds, which in this respect I take to be pretty much the same with humanity all the world over, is to make any admission of guilt. I certainly can add that, during 14 years' active practice in the criminal Courts in England, I do not remember half a dozen instances in which a real confession, once having been made, was retracted. In this country, on the contrary, the retraction follows almost invariably as a matter of course, and though I am well aware how this is sought to be explained by a suggestion of the influence brought to bear upon the confessor by other prisoners in havalat, the fact remains as an endless source of anxiety and difficulty to those who have to see that justice is properly administered: (Per Straight, Offg. C.J., in Queen-Empress v. Babu Lal  6 All 506, at pp. 542 and 543).
16. The remarks hold good today no less than they did when they were made by Straight, J. I respectfully agree with every word of what he said. In most cases of serious gravity or difficulty we are faced with the problem of retracted confession and the value to be assigned to it against the person making it or his co-accused. In a large number of such cases there is little or no reliable evidence in corroboration. To do lip service to the doctrine which requires corroboration and to accept such evidence in corroboration, for the sake of formality, as a Judge would not conscientiously believe is not in keeping with judicial integrity.
17. It has been stressed over and over again that, when a confession is made and subsequently retracted, the committing Magistrate and the Sessions Judge should enquire into all the circumstances in which it was made and those in which it was subsequently retracted. One however seldom comes across a case in which any serious effort is made cither by the Magistrate or the Sessions Judge to explain the phenomenon that a person who took every precaution of concealing his crime and suppressing evidence which could implicate him becomes so full of remorse and penitence when he comes face to face with the police that he makes a "free and voluntary confession," but subsequently, retracts it at the enquiry, the remorse and penitence, which are supposed to have acted on his mind before, ceasing to influence it in the slightest degree. While it is true that in some cases voluntary confessions are made and while it is permissible for police officers to resort to legitimate devices to obtain useful information from prisoners, it is inconceivable to me, as it was to Straight, J., that ordinarily a dacoit or murderer would make a voluntary confession. It is therefore necessary that, when a confession is first made and is subsequently retracted with allegations against the police, the Magistrate and the Sessions Judge should probe the matter for their own satisfaction. What however is done in practice is to record a confession with due formalities and subsequently to record the retraction thereof, leaving it to the accused to get over if he can the effect of the confession which stands against him in spite of retraction. It is impossible for the accused, even if he is defended, to adduce any reliable direct evidence of mal-treatment or inducement while he was in the police custody. His allegations, when put to the investgating officers, are naturally denied. But the matter should not be allowed to rest there. The Judge, with whom the responsibility lies for acting upon the confession, should satisfy himself by putting searching questions to such witnesses as had anything to do with the confession. The first question that ought to strike every Judge is "why the accused made the confession?"
18. It is very important to ascertain, from those in whose custody the accused was, the circumstances in which the question of confession first arose, how the accused expressed his willingness to be placed before the Magistrate and his readiness to make a confession. Similar questions arise as regards retraction. It is only if circumstances make it reasonable to believe that the accused voluntarily made the confession and agreed to make it before the Magistrate that an inquisitive mind can be satisfied. In the present case the statement of the invenstigating officer, made before the Sessions Judge does not betray the slightest anxiety on the part of the learned Sessions Judge to elicit any information as regards those circumstances. Such of them could be ascertained from the record failed to attract his attention. For all these reasons, I am in entire agreement with my learned colleague in holding that Nazir's confession is wholly inadmissible in evidence and should be excluded from consideration. As regards the confession of Idu the case is different so far as he is himself concerned; but I am not prepared to give much weight to it as against his co-accused. (The judgment then considered the evidence and after holding that Idu's confession was not the result of any inducement, proceeded). While I am prepared to take it into consideration under Section 30, Evidence Act, against his co-accused, I am not prepared to give much weight as against them. I have already expressed the view in an earlier part of this judgment that the evidence of the approver, coupled with the retracted confession of a co-accused but uncorroborated by other evidence, is not sufficient for conviction of the co-accused. There is an additional reason in this case, namely, that Idu's retracted confession as against his co-accused is otherwise untrustworthy. Nazir's own confession having been excluded, there is no evidence against him, except that of the approver and the retracted confession of Idu. No stolen property identified as such has been traced to his possession. As regards the appellants other than Idu, my learned brother has discussed the evidence in detail as against each and has shown that there is no evidence against any of them, except that of the approver and the retracted confession of Idu. (Then the judgment considered the evidence of the identification of accused and the stolen property and holding it unsatisfactory, proceeded.) I agree in upholding his conviction and enhancing his sentence to transportation for life which is the only proper sentence having regard to the gravity of the offence. I also agree in acquitting the rest of of the appellants.