P L D 1960 (W (PLP)

S. A. HAQ, J.‑--The 10 appellants were tried by Mr. Fazle Illahi A. Ansari, Additional Sessions Judge, Larkana, on charges under sections 120‑B, 148, 302, 326 and 449, P. P. C. read with section 149, P. P. C. Appellants Gul Hassan Wazir, Bhaji and Mahmood were further tried under section 19‑E of the Arms Act. Convictions and sentences recorded were as follows: (a) All the 10 were convicted under section 120‑B of the P. P. C. and all of them except Khairo and Shaman were sentenced to death while Khairo and Shaman were awarded transportation for life. (b) Wazir, Bhaji, Mahmood, Sheru, Himatali and Gul Hasan were further convicted under section 148, P. P. C. and each of them sentenced to 2 years' R. I. (c) The six persons just mentioned were also sentenced to death under section 302/149, P. P. C. (d) These six accused were further convicted under section 326/ 149, P. P. C. and each of them sentenced to undergo R. I. for 6 years. (e) These six persons were also convicted under section 449, P. P. C. and each of them sentenced to undergo to R. I. for 7 years. (f) Finally, Gul Hassan, Wazir, Bhaji and Mahmood were convicted under section 19‑E of the Arms Act and each of them sentenced to undergo R. I. for 3 months. 2. All of them have appealed against these convictions and sentences, and the sentences of death are also before us for confir mation under section 374, Cr. P. C. 3. Briefly, the prosecution case was this. There was enmity between Ahmadali appellant No. 10 and Illahi Bux (P. W. 15). A few days before the occurrence some mud clods were thrown into the house of Illahi Bux's son‑in‑law Allahrakhyo (P. W. 14) and the appellant Ahmad Ali was suspected of this mischief. There was a quarrel between Illahibux and Ahmadali, and during this scuffle Ahmadali's turban fell down. Ahmadali took it as a great insult and declared that Illahibux had behaved in this manner on the strength of Khanan's party and that he would take revenge. To satisfy this desire for revenge, Ahmadali hatched a conspiracy in which the other nine appellants played different parts, and in pursuance of this conspiracy an assault was carried out on the house of Khanan at about 8 or 9 p.m. on 2‑7‑56. Guns and hatchets were used and four persons were murdered, namely three brothers Khanan, Muhammad Bux and Sadar and their uncle Allahdino. A fifth man Shahan (P. W. 11), also an uncle of the three deceased brothers, was seriously injured but survived. The attack was witnessed by the women folk in the house, namely Mst. Amnat (P. W. 8) wife of Muhammad Bux deceased, Mst. Umrah (P. W. 9) wife of Khanan deceased, and Mst. Shamul (P. W. 10) wife of Sadar deceased. Besides, other relations of the deceased persons were also attracted to the spot from the neighbouring houses. They were Alinawaz (P. W. 7) who is uncle of the deceased brothers, Rajab (P. W. 13) who is a cousin, Allabrakhyo (P. W. 14) a son‑in‑law of Illahibux, and Siddik who was a brother of Alinawaz and who died during the trial. For that reason his deposition before the committing Magistrate was brought on the record of the trial under section 33 of the Evidence Act. 4. Alinawaz (P. W. 7) proceeded to inform the bead man Rasulbux (P. W. 18) about the incident and was advised to report the matter to the police. He reached the police station Damrah at 8 a. m. next day and F. I. R. Exh. 21 was recorded at his instance. The distance between the police station and the scene of offence in village Kothi Kalhoro is stated to be 15 miles. The investigation of the case was originally started by H. C. Ghulam Qadir but later taken over by Sub‑Inspector Nur Muhammad (P. W. 35). 5. The post mortem examination on the four dead bodies was performed by Dr. Ziaul Haq and his evidence shows that there were gun shot wounds on all the four which had proved fatal. In addition, he found three incised wounds on the body of Sadar. 6. The same doctor also examined the injured Shaman and found that he had two incised wounds, one on the right shoulder and the other on his left buttock. There was also a scratch on the right side of the neck. 7. All the appellants pleaded not guilty at the trial, and the general plea taken was that they had been implicated due to enmity. Four accused persons made judicial confessions during the investigation. Two of them namely Soomar and Aligohar were subsequently made approvers and have appeared as P. W. 20 and P. W. 21 respectively. The other two namely Khairo and Shaman (appellants No. 7 and 9) however retracted heir confessions during the trial. 8. The prosecution case rests on the following (1) Evidence of motive. (2) Ocular evidence. (3) Evidence of the approvers. (4) Judicial confessions of the appellants Khairo and Shaman. (5) Recoveries of incriminating articles. 9. We first take up the question of motive. In the first information report Alinawaz (P. W. 7) stated that the accused committed these murders because Ahmadali and his companions had caused, nuisance in the whole of the village and the villagers had companied against them before the Deputy Collector, Larkana. He further added that in those complaints the two deceased Allahdino and Khanan were witnesses. It will be seen that he did not mention in the F. I. R. the incident of the scuffle between appellant Ahmadali and P. W. Illahibux during which Ahmadali's turban is supposed to have fallen down. At the trial, however, Allahrakhyo (P. W. 14) Illahibux (P. W. 15) and Ramzan (P. W. 30) have deposed to this incident, in addition to Alinawaz and others. Illahibux has also produced an application Exh. 38 which was addressed by him, to the Political Agent of Sibi complaining against Ahmadali, Aligohar and Bahadur. This is dated 21st June, 1956 i.e., a few days before the incident. Mr. Pirzada, appearing for appellant Wazir, has drawn our attention to the fact that it is remarkable that this application was addressed to the Political Agent of Sibi whereas the parties resided in the Larkana District and the proper authority to be approached was the District Magistrate of Larkana. In this application Illahibux had alleged that he apprehended mischief from Ahmadali and his associates. Mr. Pirzada has argued that the incident of the falling of the turban of Ahmad Ali at the hands of P. W. Ilahi Bux is not proved on the record and in any case is too trivial to form an adequate motive for such a serious revenge as is attributed to Ahmadali in the present case. It seems to us however that whatever be the truth about the turban incident, the fact stands established on the record that there was enmity between appellant Ahmadali on one side and P. W. Illahi Bux on the other. The latter has given sufficient details in his evidence at the trial, and even Ahmadali has admitted in his statement under section 364 of the Criminal Procedure Code that Illabi Bux has involved him and others in order to save his cultivators. It is in evidence that the deceased persons were haris of Illabibux and were his party men. 10. As already stated, most of the eye‑witnesses are related inter se and to the deceased persons. We have also just mentioned that there is enmity between the party of the appellant Ahmad Ali and that of P. W. Illahi Bux which included the deceased persons and the injured Shaman. For these reasons, it is necessary to scrutinise the evidence of the eye‑witnesses with great care. 11. In the First Information Report Ali Nawaz (P. W. 7) mentioned that the assault was lodged by 12 men out of whom he identified 7 persons as follows (1) Ahmadali. (2) Ali Gohar. (3) Umedali. (4) Bago alias Abdul Rahman. (5) Rahimbux. (6) Bahadur, and (7) Khairo. Ali Gohar named by the first informant died during the trial Umedali was left off by the police after arrest, Bago was not even arrested, and finally Bahadur was also not prosecuted. While giving evidence at the trial, Ali Nawaz completely changed his position and stated that the assault was carried out by 6 persons only, and that all of them were unknown to him. The reason for this somersault appears to be that on the day of the incident the appellants Ahmedali and Rahimbux were in the Khairpur jail, having been arrested under the Excise Act. At the trial, Alinawaz tried to explain the inclusion of these names in the F. I. R. by saying that he had meant to indicate that the murders bad been arranged or instigated by Ahmadali and Rahimbux. The explanation, however, is clearly unacceptable, as the recital in the F. I. R. is definite to the effect that the persons, who raided the house and committed the murders, included Ahmad Ali and Rahimbux. It is conceded by the learned counsel appearing for the State that Ali Nawaz did falsely include the names of Ahmadali and Rahimbux as the actual assailants. It seems to us that no reliance can be placed on the evidence of Ali Nawaz, as this man has deliberately included two persons in the list of actual assailants, who were admittedly lodged in the Khairpur jail at the time of the incident. Some of the other persons, named by him, have not been prosecuted, and for that reason as well, it is clear that the story put forward by Ali Nawaz at the earliest opportunity has not been accepted. 12. The three female witnesses are all unanimous on the point that they identified the appellant Gul Hassan as being one of the assailants, and that they cried out his name at that very time. If that is correct, then there is no explanation as to why the name of Gul Hassan was not mentioned in the F. I. R. The only explanation advanced is that Ali Nawaz had already left the scene of offence before these women had time to mention the name of Gul Hassan. This explanation, however, is not correct, as, in the first instance, most of these witnesses who incrimi nate Gul Hassan by name, admitted during the commitment pro ceedings that Ali Nawaz was present at the spot when these witnesses mentioned the name of Gul Hassan, and they were duly not confronted by their previous statements in this behalf. In any case, Mst. Amnat (P. W. 8) has also admitted at the trial that the name of Gul Hassan was mentioned before Alinawaz left for the Police Station. 13. A second significant fact in this context is that Ali Nawaz admittedly reached the Police Station Dhamrah at 8 a.m. the next day, and before that he reported the matter to Rasulbux (P. W. 18) at 5 a.m. and the explanation for not reporting the matter earlier to Rasulbux or to the Police is that immediately after the incident in the earlier part of the night on 2‑7‑1956. there was a storm which prevented Ali Nawaz from travelling out. If this explanation is accepted, and delay in reporting the matter to the Police is not used against the prosecution, then it becomes clear that Alinawaz remained at the spot for several hours after the incident. Even otherwise, it seems quite natural that when four of his nephews had been murdered, and another relation seriously injured, Alinawaz would not at once rush to the Police station, without trying to ascertain as to whether any of the culprits had been identified by any of the witnesses. It seems to us, therefore, that the failure of Alinawaz to specifically mention the name of Gul Hassan in the F. I. R., clearly points to the conclusion that the appellant Gul Hassan was not identified at the spot by any of the witnesses. 14. Similar is the case with the appellant Wazir. Rajah (P. W. 13) has asserted that he identified appellant Wazir as one of the persons who was present in the street outside the house of Khanan deceased as he knew Wazir previously. It is clear that if indeed Ra jab bad identified Wazir, he was bound to mention the fact to Alinawaz before the latter proceeded to report the matter to the Police. We have already rejected the explanation that Alinawaz left immediately and, therefore, the name of Wazir could not have been communicated to him. In fact, Rajab admitted before the Committing Magistrate that Alinawaz was present at the scene when Rajab had arrived at the spot, and that Alinawaz had shouted to this witness that he was going to Rasulbux, and then to the Police Station. It seems to us that in the face of these facts, the omission of the name of Wazir from among the list of assailants is very significant, and casts a great doubt regarding the presence of Wazir at the time and place of the incident. 15. The learned Additional Advocate General has argued that the F. I. R. is not a piece of substantive evidence, and any omissions or contradictions found therein can only be used to discredit the maker thereof, and not the other witnesses. We have no quarrel with these propositions, but it is clear that the F. I. R. very valuable and important document, and if the case, set up at the very earliest moment, that is, at time of giving the first information, is completely contradicted by the witnesses appearing at the trial, a grave doubt is credited in the mind of the Court which needs to be cleared up and explained satis factorily. In the present case no satisfactory explanation has come on the record regarding the omission of the names of Gul Hassan and Wazir from the first information report, and we have already held that the first informant Alinawaz remained at the spot for several hours after the incident, during which there was ample opportunity for the witnesses identifying Gul Hassan and Wazir to inform Alinawaz of the fact. The failure of Alinawaz to mention these two names in the F. I. R., there fore, can mean either that Alinawaz has deliberately omitted the names of these two persons for reasons best known to him, or that the two names were never mentioned to him for the reason that the two men were not identified at the spot. Alinawaz is a close relation of the deceased persons, and he has no apparent reason to favour Gul Hassan and Wazir by omitting their names. The only inference, therefore, which can be drawn from the omission of their names in the F. I. R. is that they were not identified at the spot and, therefore, the case against them becomes doubtful. 16. No doubt at successive identification parades held by the Investigating Officer, the various appellants before us were identified by several eye‑witnesses, but we consider it unnecessary to enter into the details of these identifications. In the first place we find all these identification tests were held by a police officer and not by a Magistrate. It has been repeatedly observed by this Court as well as by the Supreme Court that identification parades held by Police Officers are very weak type of evidence, specially where it is not shown that the services of a Magistrate could not be made available. The identifications in this case were held at Larkana where a Magistrate could obviously have been made available without much difficulty or inconvenience. Secondly, the night of the incident was a dark one, and it is stated that the features of the culprits were observed by the witnesses in the light of a lamp which was burning in the court-yard where the murders were committed. Mr. Khalid Ishaque, the learned Additional Advocate‑General, has laid great stress on the fact that the assailants were obviously able to identify their victims, as none of the women were harmed although the fire‑arms and hatchets were used, and the women were sleeping in the same beds as their husbands. The learned counsel has, therefore, argued that similarly the witnesses must also have been able to identify the assailants who fired from a very close range, as is apparent from the fact that most of the gunshot injuries were blackened and charred. While there is no doubt that the assailants were able to pick out their victims without any mistake and none of the other inmates of the house was harmed, there is one difference between the identification by the assailants and that by the witnesses. The assailants clearly knew their victims for whom they had come, while the witnesses have deposed at the trial that except Gul Hassan and Wazir, none of the assailants were known to them. To identify a known person at night is clearly different from fixing in one's mind the features of an unknown person for the purpose of a subsequent identification. It is correct that the eye‑witnesses have no direct personal enmity with the individual appellants, but it appears to us that from the very beginning the case set up by the prosecution was not honest, and for that reason we find it very difficult, and indeed highly unsafe, to rely on the testimony of eye‑witnesses who are related to the deceased and who, as a party, are inimical to the appellants. These considerations have to be taken along with the fact that in the F. I. R. at least two persons were mentioned as participants who were admittedly in the Khairpur jail. We have, therefore, formed the conclusion that the ocular evidence in this case is not at all dependable, and cannot be acted upon, unless it is satisfactorily corroborated by other independent evidence. 17. We next have the evidence of the two approvers, namely, Soomar (P. W. 20) and Ali Gohar (P. W. 21). Soomar was arrested on 26‑7‑1956 and made a confession before a Magistrate on 27‑7‑56, while Ali Gobar was arrested on 30‑7‑56 and made a confession on 1‑8‑56. They were both made approvers on 6‑8‑56, and subsequently released on bail. We have not been able to discover the necessity for making an approver, much less two approvers, in this case. As regards the actual incident, there was already available the direct testimony of the eye‑witnesses, and the confessions of the four accused persons, including the two approvers, had already been recorded several days before pardon was tendered to these men on 6‑8‑1956. It goes without saying that ordinarily the only justification fort tendering pardon to a person who has participated in a crime is the necessity of obtaining evidence in a case where otherwise such evidence is not available. This was hardly the case here, as besides ocular evidence, judicial confessions of four persons were already on the record. In any case, if an approver was considered necessary, then it is not understood why two persons were tendered pardon and not only one, who could have given the necessary clues about the discovery of the main culprits in the case. We are, therefore, inclined to look with great suspicion on the evidence of these two approvers. Another remarkable fact about these approvers is that they were released on bail immediately after the tender of pardon on 6‑b‑1956 in contravention of the clear provisions of law contained in sub‑clause (3) of section 337, Cr. P. C. namely:-- "Such person, unless he is already on bail, shall be detained in custody until the termination of the trial." While we are not prepared to go all the way with Mr. Abdus Sattar Pirzada that the enlargement of the approvers on bail vitiates their evidence altogether, we consider that the grant of bail is certainly an element which weakens their evidence, as it seems to us to be a clear inducement or temptation which must have been offered to these two persons for becoming approvers. We also note that after the grant of pardon the two approvers were not required to make any other statement until the commencement of the inquiry. 18. Apart from these circumstances attending the grant of pardon and release on bail, we find it rather difficult to persuade ourselves to believe that the two approvers voluntarily appeared before the police on learning that they were involved in a murder case. The approvers are not shown to be directly connected with Ahmedali or his cousin Rahimbux, who are alleged to be at the bottom of these murders. The only link is stated to be that Ali Gohar was a servant of Ahmedally. The statement of Soomar makes an interesting reading. He says that Ali Gohar was his friend and he came to him at his "otak" and asked him to help in carrying out the evil designs of Ahmedali, and Soomar at once replied that he was ready to help. The only consideration, therefore, which prompted Soomar to join in this dastardy crime was leis alleged friendship with the other approver Ali Gohar. At the crucial moment Ali Gohar did not join the raiders, but Soomar remained with them till the last, and parti cipated in the actual assault. The two approvers, on their own admissions, are bad characters who have been previously concerned in several crimes. Another fact which might be mentioned at this stage is that according to document Exh. 112 (at page 176 of the paper book) the approver Soomar was in police custody on 23‑7‑56 in connection with theft case No. 85 in 1956 of Larkana town Police Station, and at the request of the police 10 days remand was granted by Mr. Abdul Wali, Sub‑Judge and F. C. M., Larkana on 24‑7‑1956. Now, according to the prosecution, Soomar first went to one Fateh Muhammad, the "`nekmard" of his village, and then voluntarily appeared before the police officer at Bismilla Hotel Larkana on 26‑7‑1956. The learned A. A. G. has suggested that the man must have been released on bail after the grant of remand by the Magistrate, but there is nothing on the record to support such a suggestion. We are, therefore, left wondering as to whether the voluntary surrender of the approver Soomar on 26‑7‑56 is not a piece of padding on the part of the investigating agency. Such being the circumstances regarding the approvers, we have no hesitation in saying that the evidence given by these two men does not inspire any confidence in our mind, and in any case even under the best of circumstances, the evidence of the approvers would require corroboration in material particulars. 19. The next link in the chain of the prosecution evidence is furnished by the judicial confessions of the appellants Khairo and Shahan. As already mentioned earlier, the two appellants have retracted their confessions and alleged that the confessions were obtained from them through torture. There is no evidence about the alleged torture but there are certain circumstances which have to be taken note of. In the first place it is in the evidence of the Magistrate, Mr. Gohar Ali Baloch who recorded the confessions, that the two confessing accused were sent back to police custody after they had made their confessions in our view, this circumstances seriously vitiates the voluntary nature of the confessions, and it is indeed a salutary practice that after the recording of confessions, the accused persons should be sent G to judicial custody, and they should in fact be informed before hand that whether they make a confession or not, they shall not be delivered back to police custody. In the present case the reverse has happened. The investigating officer has made it clear that he had requested the Resident Magistrate before hand to deliver the accused back to his custody after recording the confessions. We cannot rid ourselves of the feeling that this request of the investigating officer for the delivery to him of the two accused persons must have interfered with the voluntary nature of their confessions. In the second place, the confessions are not free from elements of exculpation. The appellant Khairo has stated that when he and other had delivered weapons to the assailants, then he was asked to run away, and after a while he heard gun reports from the house of Khanan's party and remained concealed from place to place until the police suddenly arrested him at Larkana. Thus Khairo implicates himself only as far as the offence of conspiracy is concerned, but not in the actual commission of the crime. Similar is the case with accused Shaman's confession, as he also stops short of participation in the actual assault. Thus the two confessions seem to be suffering from the following weaknesses:‑ (a) After the confessions the two accused were sent to the police custody, and this fact vitiates materially from the volun tary nature of the confessions; (b) The confessions are exculpatory in so far as the offence of murder is concerned, and (c) The confessions have been retracted. While there is no doubt that the confession of an accomplice can be taken into consideration as against the other co‑accused under section 30 of the Evidence Act, but such a confession, particularly a retracted one, needs corroboration in material parti culars. And where we are doubtful about the voluntary nature of the confessions, they can hardly be credited with any evidentiary' value at all. 20. Lastly, we have certain recoveries of incriminating articles at the instance of the approvers and some of appellants. It is stated that a pair of shoes was recovered by the investi gating officer at the instance of Himathali appellant on 27‑7‑1956 at a distance of about 15 paces from the scene of offence, where it was found stuck in the mud. These shoes were later tried on appellant Wazir and found to fit him. Now it seems to us that this particular recovery is completely worthless. The investigating officer first visited the scene of offence and on 3‑7‑1956 and he prepared an elaborate mashirnama giving the details of the various incriminating articles discovered and seized by him at the spot. It is significant that he could not notice a pair of shoes stuck in the mud at a distance of only 15 paces from the scene of offence. It is well known to every investi gating officer that shoes are an important piece of evidence in a case of the present kind, and, therefore, we are inclined to the view that the shoes were not available at or near the spot soon after the murder, and their discovery on 27‑7‑1956 is of no value whatsoever. 21. Then there is the production of a single barrel 16 bore gun by approver Soomar from his house on 26‑7‑56. This gun was sent for examination to the Ballistic Exert, Khurshid Ahmad (P. W. 12) on 18‑10‑1956. The expert did not fire any test cartridges from this gun because he considered it to be unsafe, and accordingly he applied an alternative test of preparing a mould of the breach face of the gun and comparing the charac teristics with the crime cartridges recovered from the spot and found that one of the 16 bore evidence empties had been fired from this weapon. 22. Mr. Pirzada has pointed out that the method of identi fication adopted by the Ballistic Expert in the case of the 16 bore gun produced by Soomar approver is not satisfactory, and no reliance can be placed on it for the reason that it does not take into account the element of pressure which can only be tested by firing a test cartridge from the crime weapon. Secondly, he pointed out that mass‑produced weapons will give similar identi fication marks on cartridges fired from them and for that reason it is necessary to employ several similar weapons before coming to any definite conclusion as to whether the weapon seized in the case was the one used for firing. Mr. Pirzada has referred us to the observations on pages 133, 134 and 175 of Burrad's "Identification of Fire Arms and Forensic Ballistics. We think there is force in this criticism, but more important is the delay which has taken place in this case in sending the weapon and the crime cartridges to the Ballistic Expert for testing purposes. As just stated, the gun was produced by Soomar on 26‑7‑56 but it was not despatched to the Expert until 18‑10‑56. This delay, in our view, materially detracts from the value of the evidence, provided by the identification of the gun as one of the weapons used in committing the offence. 23. A second gun was recovered at the instance of appellant Bhajhi on 28‑7‑56 and this was a 12 bore gun. Seven 12 bore empties had already been recovered from the spot on 3‑7‑56. In this case the Ballistic Expert applied the firing test and came to the conclusion that six 12 bore crime empties appeared to have been fired from the left barrel of this gun. But here again there was the same delay in sending the weapon to the Ballistic Expert, and the same criticism applies. In considering the effect of this delay, we are also influenced by the fact that in the relevant mashirnamas relating to the recovery of the empty cartridges from the spot, there is no mention at all that the empties were made into a sealed parcel. In other words, we consider that when there is no evidence that the empty cartridges or the crime weapons recovered were made into sealed parcels at the spot, and where there has been inordinate delay in sending the crime empties, and the weapons from which they were alleged to have been fired, to the Ballistic Expert for identification purposes, the value of the evidence furnished by these recoveries and identification tests is considerably weakened, and it will not be safe to place reliance on such evidence in a major crime. 24. It will thus be seen that although enmity between the parties is proved, the immediate cause which may have led to this conflagration is not thoroughly established. The evidence of the eye witnesses, or of the approvers, or of the confessing appellants, or that furnished by the recovery of certain incrimi nating article is not such as can individually form the basis of a conviction. Each of these pieces of evidence requires corro boration in material particulars, and the question is whether they can be used to corroborate each other. The learned A. A. G. has strenuously urged before us that we should use the evidence of the approvers or of the judicial confessions or of the recoveries as corroborating the ocular evidence or vice versa. 25. Mr. Khalid Ishaque relies on the following observations of their Lordships of the Supreme Court in Rafiq Ahmad v. The State (P L D 1958 S C (Pak.) 317): "The view that the confession of an accomplice does not in any circumstances furnish sufficient corroboration of the testimony of an approver overlooks the provision in section 114 of the Evidence Act that while presuming that an accomplice is unworthy of credit unless he is corroborated in material particulars the Court shall have regard to facts to be found in the illustration appended to illus. (b) in considering whether the above maxim does or does not apply to the particular case before it, the illustration, namely : "A crime is committed by several persons. A, B and C, three of the criminals are captured on the spot and kept apart from each other. Each gives an account of the crime implicating D and the accounts corroborate each other in such a manner as to render previous concert highly improbable". From this it follows that there are cases in which an account of crime given by an accused person implicating his co‑accused can be taken into consideration as corroborating the approver." It appears to the learned Additional Advocate General that these observations of their Lordships can be construed to mean that one piece of weak evidence which would not suffice by itself to form the basis of conviction can be used to corroborate another similar piece of evidence. 26. It seems to us, however, that their Lordships of the Supreme Court were dealing with a limited question in Rafiq's case, namely, whether the confession of an accomplice could furnish corroboration of the testimony of an approver, and they have based their answer on an illustration contained in the Evidence Act itself. Even then their Lordships proceeded to sound a note of caution in the following words: "But Courts should be very careful in making use of the confession of the co‑accused as corroboration of an approver's testimony, and the probability of, the confession having resulted from previous concert should be excluded before it can be used for corroborating the approver. As regards a confession which is retracted reasons given for retracting it must be found to be false before it can be acted upon. But where it is found that previous concert was highly improbable and the reasons for retracting the confession are untrue there can be no justification for declining to rely on it for purposes of corroboration." It will be seen therefore that Rafiq's case cannot be used as an authority for the general proposition that one piece of weak evidence can safely be corroborated by another piece of similarly weak evidence which in turn would require corroboration and could not form the basis of a conviction. 27. A similar question came up for consideration in Shah AU and others v. The Crown (P L D 1954 Sind 136) before a Division Bench of the former Chief Court of Sind, of which one of us (Inamullah, J.) was a member, and the conclusions were stated thus: "The question before us is whether where there are different sets of evidence and neither of which alone can be accepted without corroboration, can each in turn be taken to corroborate the other and joined together, justify the Court to act on such evidence. We are of the opinion that such evidence which requires corroboration cannot corroborate each other and there fore cannot be legal basis for conviction." In coming to this conclusion earlier decisions in Drendra Bhatta Charya and others (A I R 1927 Pat. 257) and Empress v. Jadub Das (I L R 27 Cal. 295), were followed:‑ 28. Again, the same view was affirmed in Kadir Bux v. The State (P L D 1959 Kar. 7), as will be seen from the following observations appearing therein: "We have on the record a retracted confession of Muhammad Khan as well who was a co‑accused in the case, but we are excluding it from consideration altogether although the law permits us, in section 30 of the Evidence Act, to take it into consideration. As that confession has also been retracted and stands in need of corroboration from independent sources it is not safe to use it for the purposes of corroborating another retracted confession. Corroboration must emanate from a document which is complete in itself and is not in need of support from any other quarter." A similar view was expressed in Yaru v. The State (P L D 1959 Kar. 662). 29. We are clearly of the view that we would be departing from the well‑established rule of prudence if we were to hold that one piece of weak evidence, itself requiring corroboration, can be used as corroboration of another similar piece of weak evidence. It would be highly unsafe, in fact dangerous, to base a conviction on an aggregation of such weak evidence. Applying this standard to the present case, we find that the convictions cannot be sustained against any of the appellants. The eye -witnesses are inimical, interested and otherwise undependable for the reasons already indicated in the body of this judgment. The evidence of the approvers, and the circumstances surrounding the grant of pardon to them, do not inspire, confidence and there is in fact a strong probability that their statements are the result of a previous concert. The retracted judicial confessions of the appellants Khairo and Shaman suffer from two infirmities, namely, that they are substantially exculpatory in nature and, secondly, the confessing accused persons were delivered back to police custody after their confessions were recorded. The recoveries of incriminating articles are inconclusive as there has been a considerable delay in sending the weapons of offence and the crime cartridges to the Ballistic Expert for examination, and also because in one case the test applied by the expert was far from reliable. Lastly, we have already stated that even the immediate cause which is alleged to have led to the murders is not satisfactorily established on the record. 30. For all these reasons we find it impossible to uphold the convictions of any of the appellants on any of the charges brought against them. We, therefore, accept the appeal, set aside all the convictions and sentences recorded against all the appellants and acquit them. The sentences of death are not confirmed. 31. We regret to observe that four cold‑blooded murders remain unpunished in this case, but the prosecution have them selves to thank for this unfortunate result, as from the very start a dishonest version of the incident was put forward and during the investigation a considerable amount of suspicious padding was introduced. Under such circumstances the Court cannot have any assurance that innocent persons have not been falsely implicated, and there being no means, in a case like the present, to differentiate the innocent from the guilty, the Court has no alternative but to throw out the whole case altogether. K. B. A. Appeal accepted.