1. This is a plaintiff's appeal arising out of a suit for sale on the basis of a mortgage-deed, dated the 3rd of April 1917, and presented for registration on the 24th of April 1917. The suit was contested mainly by Defendant No. 2 who held a mortgage, dated the 8th of April 1917, and presented for registration on the 11th of April 1917. The defendant pleaded that the plaintiff's mortgage deed had been antedated and was in fact subsequent to the defendant's mortgage. This point has been conclusively found against the defendant. He further denied the completion and validity of the document. The Court of first instance in a very summary judgment decreed the claim. On appeal the learned District Judge held that the plaintiff's mortgage-deed was defective for want of proper attestation. He held that it was executed in the presence of only one witness Baldeo at one place, and then it was brought to the Court compound where it was attested by the second witness Kamla. Kamla was dead and could not be produced, but Baldeo was examined. The learned Judge felt inclined to assume that Kamla must have attested the document at the request and admission of the mortgagor. He, however, did not feel inclined to re-examine Baldeo because he was of opinion that even if Kamla had signed the document on the acknowledgment of the mortgagor, there was no proper attestation in law and the document created no charge on the property. He accordingly allowed the appeal by the contesting defendant and gave the plaintiff a mere simple money decree against the mortgagor.
2. The plaintiff came up in appeal to this Court and a Bench of this Court came to the conclusion that the trial of the suit by the first Court had been unsatisfactory and that inasmuch as the pleas taken by the defendants were not quite express and clear, there was a possibility of the plaintiff having been prejudiced. This Court accordingly on the 23rd of March 1926, sent down an issue whether or not the statutory provisions with regard to attestation had been duly complied with. Parties were given the opportunity to produce additional evidence if they chose. The finding now returned by the District Judge confirms what had been anticipated on the former occasion. It is found that Baldeo actually saw the executant sign the document, and attested it at one place and later on at another place Kamla, the second witness, received an acknowledgment of the execution by the executant and then signed it.
3. There can be no doubt that under the ruling of their Lordships of the Privy Council in the case of Shamu Pattar v. Abdul Kadir Ravuthan  35 Mad. 607 the signing of the document by Kamla was not a valid attestation. It has however, been pressed before us that the effect of the new Act 27 of 1926, is to make it a valid attestation.
4. In order to ascertain whether the new Act is a declaratory Act with retrospective effect so as to affect all documents executed prior to it, as well as all litigations pending at the time when it came into force, it would be necessary to examine the state of the law as it stood prior to the passing of this Act. Section 59 of the Transfer of Property Act (IV of 1869) required that a mortgage-deed must be attested by at least two witnesses. The Legislature did not expressly define what the word 'attested' meant. It left it to be understood in its ordinary meaning. On the other hand the Indian Succession Act (X of' 1865), Section 50, expressly provided how the witnesses were to sign the document and made the receipt of a personal acknowledgment sufficient. This High Court in the case of Sheikh Ghazi v. Bhawdni Prasad (1896) A.W.N. 89 and Gangdei v. Shamsundar  26 All. 69 expressed the view that attestation in Section 59 of the Transfer of Property Act did not mean that the witnesses must actually see the executant sign the document, but that an acknowledgment by the executant was quite sufficient. On the other hand the view taken by the Calcutta High Court was that the word 'attested' meant that the witnesses must be witnesses of the actual execution. Girindra Nath Mukerji v. Bejoy Gopal Mukerji  26 Cal. 246 and Abdul Karim v. Salimun  27 Cal. 190. The Madras High Court took the same view: Shamu Pattar v. Abdul Kadir Ravuthan  31 Mad. 215. The Bombay High Court, in an earlier case, Ramji Haribhai v. Bai Parbati  27 Bom. 91 took a contrary view holding that acknowledgment was sufficient. The same Court, however, in a later case, Banu v. Laxman Rao Krishna Limage  33 Bom. 44 expressed the view that an attesting witness to a deed is a witness who has seen the deed executed, and who signs it as a witness. The Calcutta view was also followed by the Judicial Commissioner's Court in the Central. Provinces. Thus prior to 1912 acknowledgment was deemed to be sufficient by this Court, but it was not considered sufficient by the other High Courts.
5. The Madras case went up in appeal before their Lordships of the Privy Council who overruled the Allahabad cases and affirmed the Calcutta and the Madras view: Shamu Pattar v. Abdul Kadir Ravuthan  35 Mad. 607. Their Lordships' decision is a clear pronouncement that the word ''attested' in Section 59 of the Transfer of Property Act meant 'the witnessing of the actual execution of the document by the person purporting to execute it.' After this authoritative pronouncement of the highest tribunal, no doubt could remain as to the meaning of the section.
6. As, however, a contrary view had prevailed in this High Court, there were many creditors who had acted upon the rulings of this Court and had taken deeds signed by witnesses on mere acknowledgments. Such creditors were likely to be affected adversely. The Legislature recognized that, so far as these provinces were concerned, there might be a great hardship although there might be no such hardship in the other provinces. Accordingly Act 26 of 1917, called the Transfer of Property (Validating) Act, 1917, was passed. The professed object of this Act was to validate certain transfers of property made prior to the 1st of January 1915, and the Act was to extend in the first instance to the United Provinces of Agra and Oudh only. By Section 2 of the Act the Legislature validated instruments of mortgage or gift executed prior to the 1st of January 1915, although they had been signed by witnesses without seeing the executant sign them, provided that such witnesses, had before signing received from the exetants personal acknowledgments of their signatures. Section 3 tinder certain circumstances allowed restoration of claims on documents prior to the 1st of January, 1915, which had been wholly or in part dismissed, rejected or withdrawn after the 30th of July 1912, (the date of the judgment of their Lordships of the Privy Council) and before the commencement of the Validating Act.
7. The passing of this Validating Act leaves no doubt in one's mind that the Legislature recognized that the documents not signed by the witnesses in the way in which their Lordships had ruled they ought to be, were invalid and that in order to remove the hardship which undoubtedly existed in these provinces it was necessary to validate such transfers. If in the opinion of the Legislature their Lordships of the Privy Council had committed a judiciary error in wrongly interpreting Section 59 of the Transfer of Property Act, it was not necessary to 'validate transfers made prior to 1915.' All that was necessary would have been to declare that they were not at all invalid. The further fact that Act was confined in its operation to these provinces in the first instance and for a limited period makes it quite clear that the law as interpreted by the other High Courts and affirmed by the Privy Council was to be left unaltered unless the Governor-General by notification extended the operation of the Act to those provinces. It would thus seem that the effect of Act 26 of 1917 was to recognize that the ruling of their Lordships of the Privy Council laid down the correct law, but it was considered advisable that law should not be made applicable to certain transactions prior to 1915 which might have boon entered into under a misapprehension of the law induced by the rulings of this Court.
8. As stated above, this Act was confined to these provinces and limited to a fixed period. All documents from the 1st of January 1915 up to 1926, throughout the whole of India, were governed by the ruling of their Lordships of the Privy Council and could not be deemed to create a valid charge so as to pass interest in immovable property unless the documents had been signed by at least two witnesses who had actually seen the execution.
9. One may ignore the historical fact that between 1917 and 1925 two attempts were made to get the Transfer of Property Act amended, but on some ground or other both the Bills were rejected. In March 1925 a Bill which was the precursor of the present Act was introduced. One may again ignore the circumstance that there are differences between the Bill as originally introduced and the Act as passed, and one must also ignore the statement of Objects and Reasons and the report of the Select Committee on the Bill. But the previous enactment of 1917 cannot be ignored when the retrospective nature of the new enactment has to be considered.
10. If the Legislature had intended to make the Act retrospective there was nothing to prevent it from saying so expressly. There is no such express declaration. But the omission of such express declaration does not necessarily show that an Act is not retrospective. The preamble, however, undoubtedly favours the appellant. It says:
Whereas it is expedient to explain certain provisions of the Transfer of Property Act, 1882; it is hereby enacted as follows.
11. The preamble is no doubt a key to the construction of the statute, but it cannot, of course, control its provisions. Nga Hoong v. The Queen  7 M.I.A. 72. The preamble may therefore suggest what the intention might have been but one has to see whether that intention has been carried out by the language of the statute.
12. In declaratory statutes one ordinarily finds the word 'declared' and not only 'enacted.' But the omission of the word 'declared' is by no means conclusive. Similarly the title of the Act, which is called the Transfer of Property Amendment Act, need not necessarily show that the previous law is sought to be amended. One may not attach any great importance to the circumstance that the Act did not come into force immediately when it was passed, but came into force on the 25th of March 1926, when it received the assent of the Governor-General. But Section 2 does not merely provide that Section 59 of the Transfer of Property Act always had and has a particular meaning, or that it must be deemed to have that meaning. On the other hand an exhaustive and elaborate definition of 'attestation' is laid down and the whole of that definition is to be 'inserted' in Section 3 of the Transfer of Property Act. For a declaratory Act all that is necessary is to declare what the law is and not to amend the Act by introducing into one of its sections certain new words.
13. The body of the Act is thus not quite in harmony with the preamble. In order therefore to decide whether the Act must be taken to have a retrospective effect it is necessary to take into account a number of other considerations. Had it embodied a mere rule of procedure there would perhaps have been no difficulty. But the law requiring how a transfer of immovable property by mortgage can be effected is not a mere rule of evidence. The provisions are mandatory and imperative. Without registration and without proper attestation there is no transfer of immovable property at all.
14. Under a statute of 1696 it was necessary in Scotland that where a deed consists of more than one page, each page should be attested in the same manner by the grantor with his usual and accustomed form of signature. An agreement for a lease was executed in 1873, written on seven pages, of which only the last page was signed in full, the remaining pages being only initialed. The Conveyancing Act of 1874, Section 39 provided:
No deed, instrument, or writing subscribed by the grantor or maker thereof, and bearing to be attested by two witnesses subscribing, and whether relating to land or not, shall be deemed invalid or denied effect according to its legal import because of any informality of execution.
15. The question arose whether this Act had retrospective effect so as to validate the previous agreement. The House of Lords in James Gardner v. Edward A. Lucas  3 A.C. 582, held that it did not. Lord Cairns, L.C., at page 593, remarked:
If the construction contended for by the appellant is correct the consequence is that this section by words which at the outset are ambiguous, must have the effect of bringing again into existence, without limit as to time, every instrument which had failed of validity, which had failed to come into existence, by want of compliance with the formalities of the Act of 1696, and must have the effect of setting up every such instrument notwithstanding the titles and the arrangements of property which might have been made upon an assumption that those instruments were absolutely invalid.
16. Lord O'Hagan, at page 601, observed:
There are cases, English cases, and I believe Scotch cases also, which make a clear distinction between matters of procedure and matters of right, as to the operation of a statute prospectively or retrospectively. In the case before your Lordships the matter is of a right created by a deed, and unless the execution of that deed fulfills the conditions prescribed by the Legislature no right exists under it,
17. Lord Blackburn, at page 603, said:
But where the effect would be to alter a transaction already entered into, where it would be to make that valid which was previously invalid-to make an instrument which had no effect at all and from which the party was at liberty to depart as long as he pleased, binding-I think the prima facie construction of the Act is that it is not to be retrospective, and it would require strong reasons to show that is not the case.
18. The next point to consider is whether by giving the Act a retrospective effect, vested rights would be interfered with or not; for it is well settled that where vested rights are to be adversely affected the strongest presumption, in the absence of clear and express words, is that an Act is not retrospective. There may be in existence thousands of documents throughout India which though purporting to be a mortgage-deed are not a mortgage-deed for want of proper attestation, and create no charge whatsoever on any immovable property. All subsequent mortgagees have therefore a vested interest to claim priority as against such creditors and are entitled to enforce their claims in preference to them. Many of these might in fact have advanced money to borrowers after having ascertained that the previous documents executed by them create no valid encumbrance on the property at all. Could the Legislature have intended by one stroke of legislation to destroy the priority of all such subsequent transferees and bring into existence a valid transfer in favour of a previous creditor, who before the Act could not be held to have any interest in the eye of the law? In the words of Lord O'Hagan, James Gardner v. Edward A. Lucas.  3 A.C. 582:
Can we suppose that in that year (1876), no right existing, as the conditions of a valid execution had not been fulfilled, the. Legislature intervened, and by equivocal words, without any expression of clear intention, and in direct opposition, apparently to public policy and private interest, established a right that never before had any existence?.... A man who, on the faith of the construction which, as I think, must now be taken to be correct, had advanced a large sum of money on a purchase would, if the view of the appellants were adopted, be deprived of all his outlay.... The vested interests are those of a man who, acting upon the law as he understands it, and as it really turns out to be, has advanced his money or changed his position; and it cannot be said that in these circumstances such interests do not deserve protection.
19. The case before us is a case in point. Under the law, as it stood when the District Judge passed his decree, Hanuman Das, though a subsequent creditor, had the first charge on the property, and Girjanandan had no charge whatsoever. Even if on the 23rd of March 1926, instead of remanding an issue this Court had accepted what the District Judge had anticipated and which has now been found to be the fact, the lower appellate Court's decree would have been upheld. Could the Legislature have intended that because the amending Act came into force between the remand and the return of the finding, Hanuman Das is to lose his priority and Girjanandan is to acquire an interest in the property which he did not possess before? The position of the creditors in the other provinces might well be worse inasmuch as the law as interpreted by the Privy Council was in no way different from that which had been administered by the High Courts there. Are all such transferees to lose their first charge on the property because the new Act would make documents mortgage-deeds which were not understood to be such before? In view of these circumstances and the fact that the correctness of the Privy Council ruling was acted upon, if not expressly accepted in 1917, one might have some hesitation in holding that the present case is governed by the new Act.
20. The learned vakil for appellant has, however, brought to our notice the judgment in Mohammad Beja v. Kashi Upadhya A.I.R. 1923 All. 725, in which it was held that the new Act was retrospective in its effect and governs previous transaction. The judgment does not show whether the significance of the Validating Act of 1917 was pressed before the Bench. Nor does it show that the possibility of the new Act affecting vested rights was argued. The judgment is based on the view that the Act is a declaratory Act because it was an Act
to remove doubts existing as to the meaning or effect of a statute.
21. It may in the first place be pointed out that it could hardly be said that a doubt existed when the highest tribunal had made a final pronouncement and itself acted upon it in subsequent cases. It might not be, in the second place, quite correct to say that all Acts which are passed in order to 'remove doubts' do necessarily have a retrospective effect. Act 6 of 1913 was intended to declare the rights of Musalmans to make settlement, and the preamble of it stated:
Whereas doubts have arisen regarding the validity of wakfs, etc.; and whereas it is expedient to remove such doubts, it is' hereby enacted as follows.
22. Section 4 used the language:
No such wakf shall be deemed to be invalid etc.,
23. Although it is the declared policy of the Government not to alter the Muhammadan Law as to gifts and wakfs, all the Courts in India held that the said Act was not declaratory and retrospective in its effect but enacted a new rule of Mahomeadan Law. This position was assumed to be correct by their Lordships of the Privy Council in the case of Mutu K.A. Ramanandan Chettiar v. Vava Levvai Marakayar  41 All. 1. And this view was followed by this Court in the case of Naim-ul-Haq v. Muhammad Subhanullah  41 All. 1.
24. It may further be said that even declaratory Acts might not have such a wide retrospective effect as to cover all past transactions. An Act which declares a certain matter to be an offence even if it were to rectify a judicial error, could hardly be held to have retrospective effect so as to make a previous act committed at the time when the law was differently understood, a punishable offence. It might also be conceivable that although an Act is made declaratory in its nature, it will not affect people who have acquired vested rights and who would be affected adversely by it.
25. That (SIC) retrospective operation ought to be given to a statute
unless the intention of the Legislature that it should be so construed is expressed in plain and unambiguous language.
26. is made clear by the case of Young v. Adam  A.C. 469, decided by the Judicial Committee of the Privy Council. Their Lordships had in a previous case held that the power generally possessed by the Crown to dismiss a civil officer in New South Wales at pleasure was restricted by the provisions of the Civil Service Act, 1884, and that the Government had no power to dismiss a Civil Servant except on the grounds, and after the enquiry, which that statute prescribed. A subsequent Act called the Public Service Act, of 1895 was passed, Section 58 of which provided:
Nothing in this Act or in the Civil Service Act of 1884 shall be construed or held to abrogate and restrict the right of the Crown as it existed before the passing of the said Civil Service Act to dispense with the services of any person employed in the public service.
27. It was held that the later Act had not a retrospective effect. Lord Watson in delivering the judgment of the Judicial Committee remarked at page 474:
It may be true that the enactments are declaratory in form; but it does not necessarily follow that they are therefore retrospective in their operation, and were meant to apply to acts which had been completed or to interests which had vested before they became law. Neither the context of the statute, nor the terms of the clause itself, appear to their Lordships to favour that result.
28. At page 476 his Lordship remarked:
It does not seam to be very probable that the legislature should intend to extinguish, by means of retrospective enactment, rights and interests which might have already vested.
29. There is one more consideration which cannot be lost sight of. If the new Act is to be retrospective then the provision in Section 2 requiring that each of the witnesses must sign the instrument in the presence of the executant must also have a retrospective effect. With the exception of a stray remark in the case of Ramji Haribhai v. Rai Parbati  27 Bom. 91, we are not aware of any reported case in which it has been expressly held that the witnesses must actually sign the document in the presence of the executant. Even the two Allahabad cases did not go to the length of expressly holding that attestation within the meaning of Section 59 of the Transfer of Property Act meant that all that was required by Section 50 of the Indian Succession Act should be done. When their Lordships of the Privy Council sent down an issue in the case of Padardth Halwai v. Ramnarain Upadhya  37 All. 474, there was no direction to enquire whether the female executants also saw the witnesses sign the document. There may be hundreds of documents executed by pardanashin ladies, bed-ridden patients or blind persons which have been signed by the executants in the presence of the witnesses in one room and the witnesses have put their signatures in an adjoining room not seen by the executants. All such deeds of mortgage and gift would shave to be wholly invalid although there has so far been no clear reported authority indicating that this must be so. To give the section a retrospective effect might hit hard transferees under such documents.
30. In view of all these points which deserve consideration, and in view of the wide effect which The interpretation of this new Act is likely to have on all transactions since 1915 entered into throughout India, it seems desirable that there should be an authoritative pronouncement by a Full Bench. We would accordingly refer the following two points to a Full Bench for decision: (1) Whether Act 27 of 1926 is retrospective in its nature so as to apply to documents executed prior to its coming into force? (2) If so, whether it would also affect litigants in pending actions who have obtained a decree from a subordinate Court before the passing of the Act but the decree is still under appeal.
Walsh, Ag. C.J.
31. I am unfortunately unable to agree with the view taken by the majority of the Court in this case. The question referred is whether Act 27 of 1926 applies to document prior to the Act coming into force. A secondary question arises whether it applies in the hearing of an appeal from a decree passed before the Act.
32. In my opinion the questions permit of only one answer. The preamble to the Act says it is expedient to explain certain provisions of the Transfer of Property Act, 1882. It enacts that in Section 3 of that Act, after the definition of the word 'instrument' the following shall be inserted, namely, 'attested in relation to an instrument means' and then follows the interpretation to be put upon the word 'attest.'
33. I read this to be statutory instruction to every body who is bound by the statute and to the Courts, to interpret the word 'attested' in the sense provided whenever it has to be interpreted. The Act came into force on the 25th of March 1926 and from that date every Court was bound by the statute whenever it has to consider the meaning of the word 'attested' in relation to any instrument to which the Act applies. There is no language qualifying or limiting the instrument to which the Act applies. If it was intended to limit it was necessary to do it by express language having regard to the generality of the language which I have quoted.
34. It is clear that if a certain view of the law is changed by the decision of a superior authority, for example, if any decision of the Allahabad High Court is overruled in the Privy Council, pending an appeal in another case, the Allahabad High Court would have to apply the changed view in obedience to the authority of the Privy Council in spite of the fact that the decision of the lower Court was right when it was given in following the old decision., It seems to me incongruous that an appellate Court should be bound to effect this result, whatever the consequence to the parties, in the case of a decision given for the first time by a superior authority and should hold itself not so bound in the case of a statutory definition given for the first time by the legislation. With all respect I regard the considerations set out in the referring order as beside the question, having regard to the plain language and simple direction contained in the Act now under consideration.
35. The case seems to me to be analogous to that decided by the Divisional Court in England in Attorney-General v. Theobald  24 Q.B.D. 567. In that case there was previous legislation imposing stamp duty upon personal property passing under a voluntary settlement. A subsequent Act enacted that the expression 'voluntary settlement' should be construed in a particular way. It was held that the earlier Act must be read as having the meaning declared by the later Act. The duty imposed by the Court upon the statute must be discharged from the moment when the statute becomes law without reference to the date of the instrument. Where the meaning is plain we have nothing to do with the consequence
36. The arguments for and against the view that Act 27 of 1926 is retrospective in its nature are set forth at some length in the order of reference. It is, therefore, unnecessary for me to repeat them in detail.
37. The strong point in favour of the plaintiff-appellant is that the preamble of the Act contains the expression 'to explain certain provisions of the Transfer of Property. Act, 1882' and Section 2 of the Act defines what the word 'attested' means. The preamble standing by itself undoubtedly gives an indication the nothing new was being enacted, but only the meaning was being explained. But the preamble is merely a key to the construction of a statute and it cannot control its substantive provisions, which may be found to extend beyond the limits of the preamble. Nga Hoong v. The Queen  7 M.I.A. 72. The law requiring how a document is to be completed is not a mere rule of procedure, but a substantive provision, James Gardner v. Edward A. Lucas  3 A.C. 582. Without proper attestation no charge on immovable property is at all created. If the Act of 1926 is to be held retrospective in its operation so as to take effect from 1882 vested rights would be affected throughout India. There is therefore a very strong presumption against its retrospective nature unless the language employed is 'plain and unambiguous. 'The form of an Act is by no means conclusive. An enactment may be declaratory in form but may not be retrospective in its operation: Young v. Adams  A.C. 469.
38. The Act is called the Transfer of Property Amendment Act and did not come into force with effect from the date when it was passed but like many other Acts from the date when it received the assent of the Governor-General. It does not expressly say that it will be retrospective and will apply to all previous transactions. Section 2 does not merely declare what has always been the meaning of the word 'attested' but an elaborate definition is to be inserted in Section 3 of the Transfer of Property Act. That definition certainly goes much further than the ordinary meaning of 'attested' as given by their Lordships of the Privy Council: Shamu Pattar v. Abdul Kadir Ravuthan  35 Mad. 607. After the pronouncement of the highest tribunal, it could hardly be said that any doubt remained as to its meaning requiring an explanation. No such doubt could certainly remain after the Legislature itself by the enactment of Act 26 of 1917, recognized the correctness of their Lordship's ruling and found it necessary to validate transfers prior to 1915 in these provinces leaving the law intact in the other provinces and for the subsequent period. If in the opinion of the Legislature their Lordships had committed a judicial error all that was necessary was to rectify that error and not to validate certain specified transfers only.
39. If the word 'explain' were not in the preamble, every one would probably have agreed that the Act was only prospective. If the Act of 1917 were not in the way, I would certainly have held the new Act to be retrospective. One course is to ignore altogether the effect of the Act of 1917 and attach so much weight to the preamble of the new Act as to make it conclusive and to govern absolutely its substantive provisions. The other course is to admit that the Legislature itself in 1917 accepted the law as laid down by their Lordships to be correct, and then to hold that the Act in reality alters that law, and therefore goes beyond the preamble, in which case it cannot be retrospective.
40. The point itself is a difficult one and the difficulty is enhanced all the more by reason of a difference of opinion. On the whole, I am still impressed more by the doubts expressed in the referring order against the respective operation of the Act and therefore agree with my learned brothers Daniels and Boys, JJ., in answering the first question in the negative.
41. The questions referred for the decision of the Full Bench are:
(1) Whether Act 27 of 1926 is retrospective in its nature so as to apply to documents executed prior to its coming into force?
(2) If so, whether it would also affect litigants in pending actions who have obtained a decree by a subordinate Court before the passing of the Act, but the decree is still under appeal.
42. The second question only arises in the ovent of the first question being answered in the affirmative.
43. Act 27 of 1926 came into force under the provision of Section 5 of the General Clauses Act, 1897, on 25th March 1926, on which date it received the assent of the Governor-General. The document in suit was executed on 3rd April 1917, nearly nine years earlier. It was a mortgage-deed requiring attestation by two witnesses to give it validity under the provisions of Section 59 of the Transfer of Property Act. It has been found that only one of the attesting witnesses saw the executant put his mark to the deed. The other attested it on an acknowledgment by the executant of the deed of his having executed it. Both witnesses attested the deed in presence of the executant. If Act 27 of 1926 applies to the deed, this was a valid attestation. Under the law in force at the time, as laid down by the Privy Council in Shamu Pattar v. Abdul Kadir  35 Mad. 607 it was not a valid attestation.
44. The first question turns on whether Act 27 of 1926 is a purely declaratory Act or not. If an Act is purely declaratory, that is to say, if it enacts nothing new but merely explains the existing law the ordinary rule of construction is that it applies equally to transactions prior to and subsequent to its coming into force, Attorney-General v. Theobold  24 Q.B.D. 567. It is, however, necessary to look at the substance and not merely at the form. An Act may be in form an amending Act, and yet, when its operative provisions are examined it may turn out to be merely explanatory. That was the case in Attorney-General v. Theobold  24 Q.B.D. 567. On the other hand an Act may be merely declaratory in form and yet may make a definite change in the law. The Musalman Waqf Validating Act of 1913 was an Act of this kind. The preamble recites that it has been enacted in order to remove doubts which have arisen regarding the validity of certain waqfs created by Musalmans. Yet the Act has been held not only by this Court, but also by the Privy Council to be prospective and not retrospective in its operation. Such an Act comes under the ordinary rules that an Act is not retrospective in its operation unless an intention to make it so clearly appears on its face. The rule does not apply to its full extent to matters of mere procedure, but where a statute affects vested interests it is not to be construed retrospectively except where no other construction is admissible Gardner v. Lucas  3 A.C. 582 and Young v. Adams  A.C. 469. It is unnecessary to labour the point that the Act of 1926, if retrospective, does affect vested rights to a very serious extent. Rights in property have grown up all over India on the basis of the definition of attestation laid down by the Privy Council in Shamu Pattar's case  35 Mad. 607 and accepted by the Legislature in Act No. 26 of 1917 which will be unsettled and in many cases destroyed if Act 27 of 1926 is held to apply to them.
45. The present is exactly the contrary case to that which came before the Courts in Attorney-General v. Theobold  24 Q.B.D. 567. In that case the Act to be construed was amending in form but declaratory in substance. In the present case the Act is declaratory in form but amending in substance. If nothing had occurred beyond the decision of the Privy Council in Shamu Pattar's case  35 Mad. 607 it might be said that the Legislature intended to declare that decision to be incorrect. The matter did not, however, rest there. The Indian Legislature by Act 26 of 1917 accepted and embodied in an Act of the Legislature the interpretation put by the Privy Council on the word 'attest.' The Act in question is described as the Transfer of Property (Validating) Act 1917. It applied only to the United Provinces of Agra and Oudh, and only to those instruments executed prior to the 1st of January 1915. Within this limited area and this limited period of time it allowed validity to documents attested by persons who did not see the executant sign them provided the attesting witness had received an acknowledgment of execution from the executant.
46. It follows necessarily from this enactment on the principle of inclusio unius est exclusio alterious that in all other parts of India and for all documents executed after the 1st of January 1915 no attestation was valid unless both the attesting witnesses had seen the executant sign the deed. This law has now been changed by the recent Act. It does not matter that the preamble declares an intention of explaining the law or that the change is made by the insertion of a new definition. The preamble may be used to gather the intention of the Act in doubtful cases, but it can under no circumstances control the substantive provisions of the Act. If we find, as we find here, that the Act lays down a new provision inconsistent not merely with the interpretation placed on the word 'attest' by the Court but with the interpretation accepted by the Legislature itself in Act No. 26 of 1917, we are bound to hold, and we do hold, that it is not merely a declaratory Act but effects a change in the substantive law and as there is nothing in the Act from which an intention to make it retrospective can be clearly inferred it must he treated as applying only to instruments executed after its commencement.
47. Where the Legislature inserts a new definition in an existing Act the effect of which is to define a term previously undefined, it would ordinarily be held that the definition was intended to be retrospective, more particularly when the preamble of the Act inserting it declared that it was expedient to explain the law but where the new definition has the effect of repealing a previously existing and inconsistent definition then in the absence of clear words showing the contrary intention the new definition will apply from the date on which the new Act came into force and all documents executed prior to that date will be governed by the definition then in force. This is substantially the position here. The Act of 1917 did not in terms define the word 'attest' but it enacted expressly that for documents executed in the United Provinces prior to 1917 the word 'attest' did not require that both attesting witnesses should have seen the executant sign and by necessary implication that in all other parts of India and for all subsequent documents it was necessary that they should have done so. The definition so enacted has now been substantially modified and as the Legislature has not declared that the modification shall be retrospective the ordinary rule applies that the new definition governs only documents executed after it came into force.
48. My answer to the first question therefore is that Act 27 of 1926 is not retrospective so as to apply to documents executed prior to its coming into force.
49. In view of this answer the second question does not arise.
50. This is a reference to a Full Bench by two learned Judges of this Court for its decision on two question to be stated presently. As usual the reference to Full Bench has been due to a difference of opinion between the learned Judges who made the reference and two other learned Judges who decided Mohammadi Bibi v. Kashi Upadhya A.I.R. 1923 All. 725. One of the questions referred to the Full Bench is a rather difficult one. It is a matter of pure accident that while both the learned Judges who made the reference are on the Full Bench and are in a position to represent their views, none of the Judges who decided Mohammadi Bibi v. Kashi Upadhya A.I.R. 1923 All. 725 are here to represent their views of the case. I therefore, sought an opportunity of consulting one of the learned Judges who decided the aforesaid second appeal.
51. The facts which led up to the reference are very briefly these: The appellant before this Court instituted the suit out of which this appeal has arisen on foot of a mortgage bond, dated the 3rd of April 1917. The suit was contested principally by a second mortgagee who holds a mortgage, dated the 8th of April 1917. The main point of contention was that the plaintiff's mortgage was not a valid mortgage under Section 59 of the Transfer of Property Act inasmuch as both the attesting witnesses to it did not see the executant sign his name on the bond. It has been found that while one of the marginal witnesses saw the execution the other put his signature on the document on receiving a personal acknowledgment from the executant of the execution. It is contended for the second mortgagee that in the circumstances of the case the mortgage of the 3rd of April 1917 is not a valid mortgage.
52. The learned Judges who decided Mohammadi Bibi v. Kashi Upadhya A.I.R. 1923 All. 725 were of opinion that the amendment of the Transfer of Property Act made by Act 27 of 1926 would validate the mortgage in question although the Act was passed nine years after the execution of the mortgage-deed. The learned Judges who made the reference were inclined to disagree with this view and hence this reference. Of course the two learned Judges making the reference did not definitely commit themselves to the view they expressed in the order of reference and it cannot be said that there is, actually, at this moment a difference of opinion in the High Court.
53. The questions referred to the Full Bench are:
1. Whether Act 27 of 1926 is retrospective in its nature so as to apply to documents executed prior to its coming into force.
2. If so, whether it would also affect litigants in pending actions who have obtained a decree by a subordinate Court before the passing of the Act but the decree is still under appeal?
54. Point No. 1.-The Transfer of Property Act did not originally contain any defini-of the word 'attested' although it required that a deed of mortgage and a deed of gift should be attested by at least two witnesses. For want of definition of the word 'attested' there arose a difference of opinion among the High Courts is India as to its meaning. Except the Allahabad Court all other High Courts held that 'attested' meant that the two witnesses required to attest should be present when the executant signed the document and should therefore see him actually sign it. The difference of opinion was set at rest when the Privy Council decided in view of the majority of the High Courts. As the opinion of the Allahabad High Court had been entertained for a long time and as it was feared that many transactions had been entered into on faith of the view expressed in Allahabad, the Act 26 of 1917 was passed to validate such of the documents that may have been executed prior to 1st January 1915, in the United Provinces of Agra and Oudh. It was then that the legislature in a way signified its approval of the interpretation put by the Privy Council on 'attested'.
55. Then came the Act 27 of 1926, passed ostensibly with a view to explain what was meant by the word attested as used in the Transfer of Property Act.
56. The question that arises is whether this definition should have effect from the date the Act 27 of 1926 received the assent of the Governor-General, viz., 25th March, 1926, or whether it should be read as having a retrospective effect.
57. It is clear that the mere form in which an Act is passed is not conclusive as to whether its application is to be in the future or it has a retrospective effect. On the other hand, it is also recognised that ordinarily a declaratory Act is supposed to have a retrospective effect unless the contrary intention is clearly expressed in the body of the Act. The preamble therefore is only a guide to the intention of the legislature but it is not an infallible guide. The preamble in this case states that the legislature thought it expedient to explain what was meant by the word 'attested' as used in the Transfer of Property Act. When you attempt to explain a certain word already used by the legislature in an enactment you assume that the meaning which you are going to put always existed there and that you are doing nothing to alter the meaning borne by the word. The preamble therefore on the face of it professes only to tell us what the word 'attested' meant since the passing of the Act in 1882. Coupled with this avowed intention of the legislature to do nothing but to explain the meaning of an already existing word there comes the definition of the 'word' attested. In defining the word, the legislature desires that the definition should be simply 'inserted' in the Act. The effect would be as if the Act had all along contained the definition since its promulgation. This 'insertion' of the definition would again point to the view that the legislature meant only to 'explain' the meaning of the word 'attested' and did not profess to express the intention that the meaning was to be attached only to documents executed after the Act came into force and not to those that were executed before that. Further let us examine the definition itself. It runs as follows:
Attested in relation to an instrument means attested, etc....
58. The use of the word 'means' again in my opinion indicates that the legislature tells us that it wishes the word 'attested' to bear a certain meaning without any reference to the date of the passing of the Act of 1926. When you say that a particular word already used by a legislature means something you intend to say that is the meaning which the word has always borne and should be interpreted as if the meaning attached to the word since the time of its first use (in 1882). Such being therefore the grammatical construction of the several words used in the enactment of 1926, it seems to me clear that the Act is a pure declaratory one and has a retrospective effect.
59. On the other hand, it has been said that from the words 'expedient to explain' it cannot be inferred that the legislature really meant to say that they themselves believe that the word 'attested' has always borne the meaning which the legislature wanted it to bear by enacting Act 27: of 1926. I recognise the full validity of this argument. I have already indicated that by passing Act 26 of 1917 the legislature of the time accepted the interpretation put by the Privy Council on the word 'attested' and having done so thought it was necessary to validate such documents as may have been executed on faith of a different interpretation. But it seems to me to be equally clear that the legislature not only wanted that the interpretation as adopted by the Privy Council should be changed but also that it should change with effect from the date of the passing of the Transfer of Property Act. The words 'whereas it is expedient to explain' were adopted not by way of telling a falsehood to the public at large but simply because those words were calculated by use of as few words as possible to serve the purpose which the legislature had in view.
60. An enactment never contains a discussion of the reasons as to why it was passed. Sometimes it is necessary to say a few words by way of an introduction to express the intention of the Act and that is why the 'preamble' becomes necessary. The legislature did not owe any explanation to anybody as to why they were altering the law. They thought it was enough to use words which would indicate their intention. If the words used clearly indicate the intention, I think it is beyond the province of the Courts to scrutinise the motive. I fully recognise that the interpretation of the word 'attested' had been definitely settled by the legislature of 1917 in a particular way. It is also clear that in 1926 the legislature wanted to widen the interpretation and to include certain methods of attestation which were not previously recognised. In other words, the legislature did mean to alter the law. The only question is whether they meant to alter it for the future or whether they meant to alter it also with a view to affect the documents already executed before 25th March 1926. The intention of the legislature is to be gathered usually not from the history of past legislation but from the language employed in the Act itself and from that alone, where that intention is clear.
61. In the view I take of the language used in Act 27 of 1926 no room is left in my mind to doubt that the legislature meant to give the Act a retrospective effect. I have already stated that if the meaning of the legislature be clear it is not open to Courts to examine the previous law or the past history of it in order to find out the meaning of the word used. In my opinion it is clear that the legislature meant by using words of that implication that the Act should have a retrospective effect and it is not therefore necessary to pry into the reasons for this enactment. But it is not at all difficult to see why the Act should have been given a retrospective effect. In a country like India, where the number of illiterate people exceeds 90% an enactment like Sections 59, and 123 of the Transfer of Property Act is always a source of distress. People are not accustomed to technicalities like the attestation of a document and it was believed that in spite of the fast that the word attested had been interpreted in a particular way for the last 44 years hardships were still caused by that interpretation. The object therefore was to widen the sense of the word 'attested' without impairing to any great extent, the usefulness of the rule laid down in Sections 59 and 123 of the Act.
62. It is true that the giving of the retrospective effect was likely to affect some subsequent transferees by depriving them of such advantage as they might secure by establishing it in a Court of law, that a document having priority in date was not a properly drawn up one. But such advantages could hardly be called 'fair' advantages. For example, in this very case the plaintiff's mortgage was executed for consideration and in good faith. The subsequent mortgagee wants to take advantage of the fact that there is a technical flaw. If, by the passing of Act 27 of 1926, such advantage is denied to the subsequent mortgagee nobody need be sorry for that. It has been hinted that the definition of the word 'attested' not only widens the previous law but also alters it in a material way. I do not think that it has any such effect. In my opinion all the documents that may have been executed strictly in accordance with the interpretation of the word 'attested' as given to the word by the Privy Council, will continue to be valid under the present definition; they will not be affected. The words in the definition,
and each of whom has signed the instrument in she presence of the executant.
63. do not necessarily mean that the executant should see the witnesses sign. If a pardanashin lady signs a mortgage by way of executing it in the presence of two witnesses and if the witnesses then and there sign the document by way of attestation, the mortgage will be good enough within the definition of the word 'attested,' because the witnesses have signed in the presence of the executant though it may be that they have not signed in the view of her.
64. For reasons given above I am of opinion that the answer to both the questions referred to us should be given in the affirmative.
65. The main question and in view of the opinion at which I have arrived, the only question is whether Act. 27 of 1926 is retrospective.
66. The facts are fully stated in the judgment of my learned brother Mr. Justice Mukerji. Mr. Justice Sulaiman discussed the matter, very fully in our referring order, with which of course I was in agreement, and I have heard nothing during the course of the argument and subsequent discussion of this case to lead me to alter my views. Adopting therefore the reason already stated in that referring order I will confine myself as far as possible to certain further general considerations.
67. It is accepted that each Act must be judged by its own language and surrounding circumstances. As regards the permissibility of looking at these latter 1 am aware that doubt has been expressed by one of us but I shall refer later to authority establishing that it is permissible to consider some of the circumstances in which a particular Act was passed.
68. While each Act must be judged by its own language and circumstances there are certain broad principles to which I shall refer. Before, however, mentioning those I note that there are two outstanding features about this short Act; firstly, that the preamble states that in the opinion of the legislature it was 'expedient' to 'explain' certain provisions of the Transfer of Property Act, 1882, secondly it is hereby enacted as follows and then follows the insertion of a definition. That definition is to be 'inserted' in the Act of 1882 after the definition of the word 'Instrument.' I am unable to attach any particular weight to the word 'inserted.' This is a word equally applicable to an Act intended to be retrospective or to an Act which may be in accordance with ordinary presumption assumed to be prospective. Every amending Act contains the expressions 'shall be inserted' and 'shall be added'; sometimes the one and sometimes the other phrase is used, appropriately to the nature of the alteration.
69. Similarly I am unable to attach any particular force to the word 'means.' It is equally used in all kinds of Acts in reference to definition; whether those definitions are in an original or in an amending Act; it is just the ordinary word appropriate to a definition.
70. Nor am I able to attach any weight to the words 'any instrument.' They: are capable, standing by themselves, of including any instrument executed in the past as well as to those to be executed in the future or of being confined to instruments executed in the past or of being confined to instruments executed in the future according to the context and the nature of the Act. To assume that those words have any one or other of these three meanings appears to me to assume the answer to the question which we have to decide and to wholly deny effect to all the approved broad principles of interpretation of statutes.
71. Judging then the operative clause of the Act by itself and as it stands, I can find nothing to differentiate it from any similar clause that might be found in any Act appearing on the Statute Book for the first time.
72. Regarding the Act as a whole the single feature which suggests that the Act is retrospective and the only fact which has given rise to doubt is to be found in the existence in the preamble of the words 'to explain.' In reference to this two generally accepted propositions may be stated. Firstly, the preamble can only affect interpretation of an operative clause where that operative clause is in itself of doubtful import. Secondly, the form of an Act is not conclusive in determining the question we have to decide. This point has been dealt with by my brother Mr. Justice Daniels with whose expression of his views I entirely concur. I would add here some comments on the case with which he deals: Attorney-General v. Theobold  24 Q.B.D. 567. In that case Pollock, B. after pointing out that each case must however depend upon the language of the particular statute and to some extent on the subject matter to which the statute applies' added these words' voluntary settlement having given rise to doubts the Legislature passed an Act etc.' I shall note later on that at the time of passing Act 27 of 1926 there were no doubts of any sort whatever existing in the mind of anybody as to what the current authoritative interpretation of the law then was. These considerations together with those adverted to by my brother Mr. Justice Daniels suggest to my mind that this case is not authority for the proposition that the Act we are considering is not retrospective but rather the converse.
73. It was suggested during the course of the argument that we are as much bound by a declaration of the legislature affecting past transactions as we are bound by a judgment in regard to such transactions of their Lordships of the Privy Council. That is of course a proposition beyond dispute, but it again appears to me to assume the answer to the very question which we have to decide. If the statute is retrospective there can be no question but that we are bound to give effect to it. But we have to determine this very question whether it is or it is not retrospective. Herein, lies a difference between statutes and the judgments of a superior tribunal. The latter are never confined to declaring what the law is to be in the future. The effect is always to declare what the law now is and has been in the past; in other words, such judgments are always declaratory and retrospective as well as prospective.
74. It would seem to be clear that none of the general tests which have been accepted in aid of the determination of the question whether an Act is retrospective or not is by itself conclusive. But that does not mean, of course, that such tests are of no value and I will proceed to refer to some of them.
75. That the form of an Act is not conclusive is clear not only from the case already referred to, Attorney-General v. Theobold  24 Q.B.D. 567 and other cases, but was very specifically asserted in Young v. Adams  A.C. 469 where Lord Watson said:
It may be true that the enactments are declaratory in form; but it does not necessarily follow that they are therefore retrospective in their operation, and were meant to apply to acts which had been completed or to interests which had vested before they became law.
76. It is conceded on all sides that the initial presumption is that an Act is not retrospective. In Modern Railway Company v. Pye  30 L.J.C.P. 314 Early, C.J., said:
Those whose duty if is to administer the law very properly guard against giving to an Act of Parliament a retrospective operation unless the intention of the Legislature that it should be so construed is expressed in clear, plain and unambiguous language.
77. It is unnecessary to multiply authority for this proposition but it may be noted that it was applied by their Lordships of the Privy Council to whose authority we are particularly subject, in Mohammad Abdus Samad v. Qurban Hussain  26 All. 119 where they said:
It is not, however, in accordance with sound principles of interpreting statutes to give them a retrospective effect.
78. The only legitimate inference is that, if an Act is not clearly retrospective, if there is any doubt upon the point, the intention of the Legislature must be taken to have been that it should not be retrospective.
79. As to what is a declaratory Act, the only definition that I have been able to find is in the most recent authority on the interpretation of statute law, Craies' 'Treatise on Statute Law,' 3rd Edition, 1922, at page 60. In modern purposes a declaratory Act may be defined as an Act passed to remove doubt existing as to the common law or the meaning or effect of any statute. Such Acts are usually held to be retrospective. Without suggesting that no Act can be retrospective unless it comes within this definition, it is at least worthy of note that in the present case if there is one thing more clear than another it is that there could have been no doubts whatever in the mind of the Legislature when it passed Act 27 of 1926 as to what the then existing law was and that the law had been universally understood in the same sense for the previous nine years. It is unnecessary for me to deal in detail with the previous history of this matter. It is set out in full in the referring order of Mr. Justice Sulaiman and myself and mention of it is to be found in the judgment of my brother Mr. Justice Daniels. By Act 26 of 1917 the Legislature made it apparent beyond all doubt that it accepted the view of the law laid down by their Lordships of the Privy Council and from that date onwards that was the unquestioned law.
80. It is, however, necessary in this connexion to note a suggestion which was made during the argument but which was then not further pressed. It was hinted that if is not open to us to consider the previous history of the matter. It may be conceded at once that it is not open to take into consideration the statements of objects and reasons, the report of the Select Committee or the proceedings in Council prefatory to an Act becoming law; but there is a great distinction between such proceedings and the previous state of the law itself and the prior Acts of the Legislature. It is the universal practice when interpreting a new Act; any clause of which gives rise to any difficulty to consider the Act whose place it takes, and I am not aware that the correctness of this practice has ever been challenged. As long ago as 1882 in Gopal Pande v. Parsottam, Das [18830 5 All. 121 Mr. Justice Mahmood, after excluding from consideration the statements of objects and reasons, etc., said:
We are no doubt at liberty to consider the general state of the law which prevailed in pari materia prior to the enactment of any statute under consideration.
81. In Pardo v. Bingham  4 Ch. 735 Lord Hatherly said:
The question is whether on general principles the statute ought in this particular section to be held to operate retrospectively, the general rule of law undoubtedly being, that except there be a clear indication either from the subject-matter or from the wording of a statute, the statute is not to receive a retrospective construction.... In fact, we must look at the general scope and purview of the statute, and the remedy sought to be applied, and consider what was the former state of the law, and what it was the Legislature contemplated.
82. Further, in the absence of anything decisive, it is undesirable to give a retrospective effect to an Act where to do so would be to alter the law applicable to a claim in litigation at the time when the Act was passed: see Leeds and County Bank v. Walker  11 Q.B.D. 84, Maxwell in the 6th Edition of his 'Interpretation of Statutes' at page 394 remarks:
In general, when the law is altered pending as action, the rights of parties are decided according to the law as it existed when the action was begun unless the new statute shows a clear intention to vary such rights.
83. That an Act is not to be interpreted, in the absence of an unambiguous direction otherwise, so as to affect vested rights, has been dealt with in the referring order of Mr. Justice Sulaiman and myself and in the judgment of Mr. Justice Daniels. Authority for the proposition is to be found passim in all the Treatises on the interpretation of statutes and it was further enunciated by their Lordships of the Privy Council in Muhammad Abdus Samad v. Qurban Husain  26 All. 119.
84. Again, it is a general proposition that an Act is not to be considered retrospectively so as to affect a transaction already completed. In Gardner v. Lucas  3 A.C. 582. Lord Blackburn said:
Where the effect would be to alter a transaction already entered into, where it would be to make that valid which was previously invalid, to make an instrument which had no effect at all, and from which the party was at liberty to depart as long as he pleased, binding.... I think the prima facie construction of the Act is that it is not to be retrospective, and it would require strong reasons to show that it is not the case.
85. How determined the Courts have always been to refuse retrospective effect unless it was beyond a shadow of doubt on a consideration of the whole Act and all its circumstances that a retrospective effect was given, is clear from the form of Act 6 of 1913 (The Wakf Validating Act) and the cases refusing to give to that Act a retrospective effect: of. Naim-ul-Haq v. Muhammad Subhan-Ullah  41 All. 1, where Piggott, J., said:
If the Legislature intended to give validity to transfers of property to an unascertained number of past transactions which had no such effect at the time when they were executed, I should have expected it to do so in vary dear terms.
86. Here also we are asked to hold that the Legislature intended to give validity to transfers of property to an unascertained number of past transactions which had no such effect at the time when they were executed. Has the Legislature in this case exhibited any such intention in very clear terms? The terms of the Waqf Validating Act were infinitely more clear. The Legislature had deliberately chosen as full a declaratory form as could well be devised, but reading the Act as a whole this Court refused to give effect to that form. In the present case the Legislature, if it intended the Act to have a retrospective effect, would appear to have carefully avoided expressing that intention in unmistakable terms. It recited no doubts, it recited no difficulties, it declared nothing and it refrained from anything but the ambiguous hint that the Act might have a retrospective effect. We have considered some of the general propositions restricting interpretations in favour of a retrospective effect. The converse proposition is thus stated by Craies:
An Act is retrospective (1) when it is expressly enacted that an enactment shall be retrospective, i, e., where such words occur as 'shall operate retrospectively', (2) where it is a necessary implication from the 'language employed that the Legislature intended a retrospective operation.
87. Emphasis was laid on, the strength of the implication required by the Judicial Committee in the case Colonial Sugar Refining Co. v. Irving  A.C. 369, where, in interpreting, the Judiciary Act, the Committed said that it.' is not retrospective by express enactment or by necessary intendment.'
88. In giving weight to the above propositions I would make it clear that I appreciate that no one of them is conclusive by itself, for instance, it is clear that every Act taking effect retrospectively can be said in one sense or another to affect rights of some sort; and similarly with the other propositions in particular cases they may be open to exception. But where I find each and all of the general propositions when applied to this case as well as all the considerations more particularly concerning this Act with the single exception of the words 'to explain' in the preamble all pointing in one direction, I find myself impelled to hold that I cannot give to this Act a retrospective effect.
89. To turn to other considerations which more particularly concern this individual Act. Did it or did it not 'alter the law?' I did not hear it argued at the Bar, nor so far as I am informed do either of my brother Judges consider that the law was not 'altered.' That the law has been 'altered,' that it has in fact been very radically altered appears to me a proposition beyond challenge.' The law as laid down by their Lordships of the Privy Council and which remained undisturbed for nine years was clear and unmistakable. The law as laid down by the Act under consideration is equally unmistakably different and different in a very important and far reaching particular, If it cannot be denied, and in fact is not denied, that the law has been 'altered,' how can it be contended that the present Act only 'declares what the law has always been.' The two propositions are a contradiction in terms. I have referred to the far-reaching effect of this alteration. To give, the Act a retrospective effect would mean to alter the law as regards all transactions which have taken place throughout India since 1882 with the exception of those that may be barred by limitation.
90. In conclusion I would say that a prolonged study of all the authorities to which I have been able to obtain access has failed to furnish me with even one single parallel case where an Act altering the law which has been in unquestioned force for nine years past has been given a retrospective effect.
91. I do not lose sight of the fact that the words 'to explain' in the preamble indicate possibly some obscure intention in the mind of the draftsman, but for the reasons I have given I find myself in no doubt as to the conclusion at which I should arrive, that is, that where the Legislature, if it so desired, could have and ought to have expressed its intention beyond doubt, I have no right to interpret this Act, ignoring the ordinary presumption, so as to give it a retrospective effect.
92. In accordance with the opinions of the majority of Judges composing the Full Bench the answer to the first question referred is that Act 27 of 1926 is not retrospective in its nature so as to apply to documents executed prior to its coming into force. In view of this answer the second question referred does not arise. (The decision of the original Bench on receipt of the opinion of the Full Bench, was as follows). In view of the opinion of the majority of the Full Bench that the new Act 27 of 1926 is not retrospective, it is clear that the deed relied upon by the plaintiff was not a valid mortgage-deed inasmuch as it had not been duly attested as required by the ruling in Shamu Pattar v. Abdul Kadir  35 Mad. 607. The appeal is accordingly dismissed. As however the findings of the lower appellate Court were not satisfactory and the question ultimately raised in appeal was an intricate question of law, we direct the parties to bear their own costs of this appeal.