This is a case stated by Pring J. for the opinion of this Court under sec. 38 of the Commonwealth Income Tax Assessment Act 1915. Albert Henry Nathan appealed to the Supreme Court of New South Wales against the decision of the Commissioner, who assessed the appellant in respect of three sums representing portions of dividends received from three companies. Those sums are £6,222 received from S. Hoffnung & Co. Ltd., £30 from the Bank of Australasia and £28 from Farmer & Co. Ltd. All those companies are incorporated under the English law, and their registered offices and central management and control are in England. They carry on business and make profits in Australia, and also carry on business and make profits in England and elsewhere outside the Commonwealth. The sums in respect of which the appellant was assessed are attributable to the profits derived by each of the companies from that part of their respective businesses which is carried on in Australia. They were, however, paid to him in England, and from the argument we understand that the dividends were declared in England also. The question of law arising in the appeal is "whether such sums were rightly or wrongly included in such assessment?"

The contention of the appellant is that the sums in question were not derived directly or indirectly by him from any source within Australia. The true "source" of each of the dividends, says the appellant, is situated in England, where the registered office and central management are found and where the dividends are declared and paid. The Commissioner, on the other hand, contends that the "source" of the dividends is in Australia, where the profits out of which they are paid are made by the company.

The foundation of the matter is found in the Income Tax Acts 1915 (Nos. 41 and 48 of 1915), that is, the Acts imposing the tax. Sec. 3 says "Income tax is imposed at the rates declared in this Act." Sec. 4, for the purpose of applying differential rates, divides the "income" into two classes, viz., (1) "income derived from personal exertion" and (2) "income derived from property," and appropriates a separate schedule to each. Specific provisions are made with respect to individuals who have incomes of both classes, and with respect to companies. But in order to understand what is meant by "income derived from personal exertion" and "income derived from property," we have to regard sec. 2, which enacts that "the Income Tax Assessment Act 1915 shall be incorporated and read as one with this Act." The Assessment Act 1915 does not give a separate definition of "income." But it defines several expressions in sec. 3. "Income from personal exertion" and "income derived by any person from personal exertion" are defined as meaning "income derived in Australia" consisting of earnings and a number of other matters enumerated. "Income from property" and "income derived from property" are defined as meaning "all income derived in Australia and not derived from personal exertion." Consequently, "income" needs no definition: it comprises the two classes, which are defined and which are mutually exclusive and are exhaustive.

The question is what income is meant by the Income Tax Assessment Act to be income "derived in Australia"; for nothing else is taxed. It is trite law that, in order to ascertain the intention of the Legislature in one part of the Act, the proper course is to read the whole instrument. Before doing so, we have to make sure what the whole instrument consists of. When the Income Tax Act No. 41 was passed, 13th September 1915, the Assessment Act passed on the same day was No. 34. Afterwards, on 15th November 1915, an amending Assessment Act was passed, No. 47. But what is extremely material is the fact that by sec. 11 of the amending Act it is enacted: "This Act shall be deemed to have commenced on the same day as the Principal Act." We have therefore to deem that both were in existence on 13th September 1915, and that the Assessment Act stood on that day as it was afterwards amended by the later Act. So reading the Act, we turn to various sections for enlightenment as to what income the Legislature has intended to tax.

Sec. 3 defines "absentee" as a person who does not reside in Australia, and includes others who need not be further referred to. This definition, combined with the frame of the taxing Act, indicates that the fact on which territorial jurisdiction was founded was not the personal presence of the taxpayer in Australia but the local situation of the source of income. This is confirmed by sec. 10, which provides that income tax shall be levied and paid upon the taxable income (that is, by sec. 3, the income after permissible deductions have been made) "derived directly or indirectly by every taxpayer from sources within Australia." The words "directly or indirectly" are of great importance. Lovell & Christmas Ltd. v. Commissioner of Taxes[1] was decided on a New Zealand Statute where the material words were simply "derived from New Zealand," and the Privy Council held[2] that they meant "directly derived"; Grainger & Son v. Gough[3] was followed, and there, at p. 341, Lord Watson used the word "directly" to denote the same idea. Therefore, while income is taxable, though only indirectly derived from an Australian source, it is clear that income not derived either directly or indirectly from some source in Australia is free from taxation.

But still the question remains: Is the "source" of the appellant's dividends "within Australia"? We must look further into the Act. This cardinal fact presents itself at the threshold: when the Legislature divides all income into income derived from (1) "personal exertion" and (2) "property" it uses language which indicates that it regards these two expressions to represent the two general "sources" of income. Particular sources - such as earnings, &c. - fall within the general source denominated "personal exertion," and all other particular sources fall within the general source denoted "property." That the Legislature itself regards these two expressions as representing the general "sources" dealt with by the Act is demonstrated by the proviso to sub-sec. 2 of sec. 18, a proviso added by the November Act. That proviso says that "if the income from either source does not amount to the sum to be deducted from that source, the balance of the sum to be deducted may be deducted from the income from the other source." The "personal exertion" source is exhaustively defined, and embraces various species of income the result of personal exertion - or labour - in Australia. The "property" source by parity of reasoning is intended to be some means of production in Australia not included in personal exertion - something which represents, so to speak, the capital fund which produces the income.

The Legislature in using the word "source" meant, not a legal concept, but something which a practical man would regard as a real source of income. Legal concepts must, of course, enter into the question when we have to consider to whom a given source belongs. But the ascertainment of the actual source of a given income is a practical, hard matter of fact. The Act, on examination, so treats it. For instance, sec. 14 enumerates certain matters which it enacts "the income of any person shall include." On inspection, these matters will be found to consist of items which might otherwise be, to some extent at all events, the subject of doubt and controversy as to whether they fell within the undefined generic term "income," as contrasted with capital.

For instance, take "profits added to the capital"; but for that express provision it might have been contended that such profits were not income at all. Par. (b) as it originally stood was in this form: "dividends, interest, profits, or bonus credited or paid to any member, shareholder, or debenture-holder of a company, but not including a reversionary bonus issued on a policy of life insurance: Provided that where a company distributes to its members or shareholders any undistributed income accumulated prior to the commencement of this Act the sum so received by the member or shareholder shall not be included as part of his income." That paragraph, so standing, would have operated so as to apply to a company in Australia having its registered office here but carrying on business operations solely outside the Commonwealth, say, in the South Seas or New Zealand or England, and crediting or paying dividends to its shareholders here. Did the Legislature then mean to tax income that resulted from business operations entirely outside Australia, and did it so intend because it regarded the presence of the head office and the declaration of dividend to be a "source" within the meaning of sec. 10, and the income therefore to be "derived in Australia" within the meaning of sec. 3? It appears to us that when Parliament made the amendments referred to (Act No. 47) in November, and declared them retrospective to the date of the Principal Act, the amendments amounted to a legislative declaration of what was always intended by par. (b); and we have now to see what that intention was.

Par. (b) was amended by making two additions. First, the "company" crediting or paying the dividend, &c., was limited to "a company which derives income from a source in Australia," or "a company which is a shareholder in a company which derives income from a source in Australia." Next, it was "provided further that where a company derives income from a source in Australia and from a source outside Australia a taxpayer shall only be taxable on so much of the dividend as bears to the whole dividend the same proportion that the profits derived by the company from a source in Australia bears to the total profits of the company." Those additions seem to us to indicate the intent of the Legislature in a most unmistakable manner.

In the first place, the exclusion from all liability to tax in respect of dividends, though received in Australia, from an Australian company, if that company itself did not derive income from a source in Australia, makes it manifest that the Legislature did not regard dividends as derived from a source in Australia merely because a company's head office was situated in, and its operations directed from, Australia, and the declaration of dividend took place here. Next, if the company derived income from a source in Australia and also derived income from a source outside Australia, the place of head office and directorate and declaration of dividends again did not govern the matter, but the real source of production of the dividend, viz., the company's actual operations should govern to the extent that they so contributed. But, further, the Legislature expressly includes another company, that is, not the company which itself carries on the operations from which directly or indirectly the profits distributed are derived, but a company holding shares in that operating company. The inclusion of the second company does not indicate that the shareholding company is one that does not derive income from a source in Australia in contradistinction to the operating company that does, for such an indication would be at total variance with the scheme and the very words of the Act. On such a hypothesis - that is, that the fund from which the individual's dividend came is not derived from an Australian source - the dividend itself could not be taxable income under sec. 10 without defeating the object of the amendments. But, as that is a reductio ad absurdum, it is plain that the inclusion of the second company is for a purpose totally different from that of drawing a distinction between a company deriving income from a source in Australia and one not so deriving it. And, if that be so, the only other purpose is to see that by no process of evasion, such as the interposition of a shareholding company, can the individual deriving income by way of dividend, &c., from the operating company escape the effect of par. (b). It shows that where a dividend is received from a company that itself by operations in Australia makes profits, or from a company that by virtue of its shareholding position receives some portion of those profits, and again distributes them among its own shareholders, the dividends so ultimately received are nevertheless to be considered, subject to the second addition by amendment, as "derived directly or indirectly from a source within Australia." That second addition, which is for the benefit of the taxpayer, treats the dividend from either the first or operating company, or the second or shareholding company, on the same footing, and provides that where the company's income is not wholly derived from sources in Australia the dividend is to be taxable only in the same proportion as the company's profits are Australian. Further examination of the Act does not weaken, but, by consistency, supports this view. For instance, we may refer to the complementary provision in par. 1 of sec. 16, dealing with the taxation of companies and allowing income distributed to be deducted.

Mr. Campbell strongly pressed upon us the view that however clearly the company's source of income was Australian, the appellant's was not. And the argument was that inasmuch as the incorporation of the company was English and its registered office and board of control and actual control were English, and the declaration of dividend was made in England, therefore the whole right of the appellant to the dividend originated in England, and could not be said to be derived directly or indirectly from any source in Australia.

Before addressing ourselves to the argument as it affects this case particularly, it is necessary to draw attention to the fact that the word "company" in the Act is not confined to corporations. The definition clause, sec. 3, defines (subject to contrary intention appearing) the word "company" as including "all bodies or associations corporate or unincorporate." The word "company," then, in par. (b) of sec. 14 includes an unincorporated association. Indeed, as Lord Wrenbury (then Buckley J.) says in In re Stanley; Tennant v. Stanley[4]: "The word company has no strictly technical meaning. It involves, I think, two ideas - namely, first, that the association is of persons so numerous as not to be aptly described as a firm; and secondly, that the consent of all the other members is not required to the transfer of a member's interest. It may, but in my opinion here it does not, include an incorporated company." See also per James L.J. in Smith v. Anderson[5]. In the Commonwealth Act it does include an unincorporated company. Consequently, whatever is the intention of the Legislature with regard to a "company" that is unincorporated, that must be the intention with regard to a "company" that is incorporated. The source of the member's dividend in each case is placed on the same footing. It could hardly be supposed that the member could escape if the companies here were unincorporated. That in itself is a most serious, if not decisive, obstacle in the appellant's path. But, lest any suggestion should present itself that notwithstanding the generality of the provision a distinction must be made where the company is incorporated, we shall consider the case from that standpoint also.

It is quite true that the appellant could not assert any title to a farthing of the dividend he received until it was declared. Until that moment it was part of the company's assets, although as profits the whole fund from which it came was available for distribution if the company thought fit to distribute it. But that, though true, is not the complete truth. Moulton L.J., in Gramophone and Typewriter Ltd. v. Stanley[6], states the position in a few words. He says: - "The profits of the corporation are not profits of any business carried on by him in a foreign country, because the individual corporator does not carry on the business of the corporation. He is only entitled to the profits of that business to a certain extent, fixed and ascertained in a certain way depending on the constitution of the corporation and his holding in it." If the argument of the appellant is sound and sufficient, then if an English company is formed for working gold, silver or copper mines in Australia, with its registered office and directorate in London, and draws every penny of its wealth from this Commonwealth, the members among whom it is distributed do not derive in any degree from an Australian source, but wholly and exclusively from an English source. That strikes one as a strange and unnatural conclusion. Is there any doctrine of law compelling its acceptance? We are not aware of any. The argument treats the income of the individual as necessarily springing from the one source, an indivisible source, either the place of declaration of dividends or the locale of the company, its principal residence, where it is said every individual corporator's "share" is situated. But that, in the first place, is quite contrary not merely to the language and scheme of the Act, as we have quoted it, but also to the general view of the subject as expressed by the Court of Appeal in Gilbertson v. Fergusson[7]. The Imperial Ottoman Bank was a Turkish corporation, and carried on business at Constantinople, London, Paris and elsewhere. It resided at Constantinople. It made profits in London, which was only a branch. It also made profits in Turkey. Dividends were declared at general meetings, which were held in London[8]. The amounts necessary to pay English shareholders were held by the London agency. The question was as to the liability of the agency itself (1) for the English profits and (2) for the amount of dividends declared and payable to individual shareholders in England. The basis of the second claim was that the individuals were entitled to the dividends and that the agency were intrusted with the payment of them. Brett L.J.[9] recognized that from a general standpoint - that is, within the Bank - the Bank's profit was "one profit and one profit only, and which is to be divided in dividends." "But," said the Lord Justice, "when one comes to apply the income tax to it the Statute obliges one to divide that profit into two. Each person resident in England, who is to receive a dividend out of that profit, receives part of such dividend in respect of profit arising from business carried on in England, and part in respect of profit arising from business carried on in Turkey." Cotton L.J. says[10]: - "When a corporation carries on its business in more than one place, the dividends are not a share of the profits arising on the transactions in any one place, but of the profits made by the entire business of the corporation, and unless for any purpose it is necessary to analyse the source from which the dividends arise, it must be taken that the dividends are not paid out of any particular fund, but out of the sum which on the whole transactions of the corporation is the profit during the year." The Lord Justice[11] proceeded to say what is very important to remember is that sec. 10 of the Act 16 & 17 Vict. c. 34, under which the question arose, was not a clause imposing on the Bank, or the English agency on behalf of the Bank, a liability to pay duty in respect of profits coming to them, but was "a mode of collecting the duty which would be chargeable on the persons who are to receive the dividends." So the decision is very much in point here. All through the judgment we find that the Lord Justice speaks of the dividends being "in respect of profits," and the Court held they were divisible according to their source. That is quite inconsistent with the argument that the place where the profits are earned by the company cannot be considered the place of the source of the shareholder's dividend, and is quite opposed to the view that under the English law, at all events, it is the place of the "share" or of the declaration of dividend that constitutes the locality of the source. We know of nothing that differentiates the English law from the Commonwealth law in this respect.

It is clear from Mitchell v. Egyptian Hotels Ltd.[12] that, in determining where the business of a company is carried on, the factors relied on by the appellant here are not exclusive. Doubtless, it leaves the question still open as to what is the source of the shareholder's income, but the real meaning of the case referred to, and others cited in it, is that the whole question is a question of fact to be determined on practical grounds. The same underlying principle should be applied here.

We had pressed upon us the case of In re Chalmers[13]. That was a decision under the New South Wales Income Tax (Management) Act 1912. By a majority of two Judges to one it was held that the source of the shareholder's income is his "share," and the locale of the share must be where the company is, and, if the place of incorporation and of central management and control be the same, that definitely determines the place of the shareholder's source of income. Harvey J. dissented, and thought that the dividend of a shareholder depends on the place where the company's profit is derived. He cited Gilbertson v. Fergusson[14]. We are not called upon to determine how far Chalmer's Case is a correct decision on the law of New South Wales, but we are certainly not prepared to accept it as a proper guide upon the Commonwealth Act - one reason, though we do not say it is the only reason, being that in the New South Wales Act then interpreted "company" is defined to mean "a company duly incorporated under any law."

Is it true that the "share" is the source of the appellant's dividend, within the contemplation of the Federal Parliament? It cannot be consistently with sec. 14 (b). If it were, then, in the case of an Australian company drawing its profits from a source out of Australia, the shareholder would be liable for tax on his dividend as income derived from a source in Australia. But that is excluded. Therefore that is not the test. Nor are we aware of any legal principle which compels us to say that the "share" is the source. What is a "share"? A share in English law is a share in the capital of the company, as it is also in New South Wales (Companies Act 1899, sec. 7 (e)). Lord Haldane L.C., in Will v. United Lankat Plantations Co. Ltd.[15], thus expresses the position: - "A shareholder comes to the company and says, I wish to contract with you for a share in your capital and so to become a shareholder. He advances his money and the terms are contained in the bargain that is made between him and the company on the issue of the share to him." The bargain differs in terms according to the circumstances, but the "share" is a share in the capital of the company. Lord Cairns, in In re Suburban Hotel Co.[16], calls it "a stake in the company." The share in the capital is not the "source," but the measure of the dividend he is to receive; the rights and liabilities created as incident to the holding of that share, including the right to receive a dividend, depend on the statutory provisions of the Act of Parliament, and subject to those provisions depend on the bargain made by the parties. The whole bundle of rights and liabilities incident to holding a share is sometimes referred to as the "share," but if the term is used in this larger sense, it is of no material assistance in determining the question before us. It is certainly true that the individual corporator neither owns the corporate property nor carries on the corporate affairs. Its profits are not his profits (per Lord Parker in Daimlor Co. Ltd. v. Continental Tyre and Rubber Co. (Great Britain) Ltd.[17]). Nevertheless, that learned Lord[18] did not agree that for every purpose "an impassable line is drawn between the one person and the others." We need not go so far as that; for, if we inquire as to the relative rights of the corporation and its members with respect to the profits of the company, we shall find that by virtue of the social compact the members are not to be regarded as entire strangers. They are not like the servants of the company or ordinary creditors, who stand outside the domain of interest in the company's property and affairs. In Bligh v. Brent[19] Alderson B., speaking of the subscription of capital to corporation purposes, says "the purpose of all this is the obtaining a clear surplus profit from the use and disposal of this capital for the individual contributors." In In re Suburban Hotel Co.[20] Lord Cairns L.J. speaks of the business being "managed in a way that will be profitable for every person." Lord Cairns applied to the corporation the reasoning of Lord Hatherley in Jennings v. Baddeley[21], where in relation to ordinary partnership that learned Judge (then Vice-Chancellor) said: "Every partnership is entered into by the partners with the view of deriving profit from the concern." In Smith v. Anderson[22] Cotton L.J. said: "Most persons when they invest their money do it for the purpose of profit, that is to say, they expect to get a profit in the shape of dividends." The principle has been very recently recognized by the Court of Appeal in In re Dawson; Pattisson v. Bathurst[23].

When the company has made its profits, though no individual corporator can lay claim to any portion of them, every corporator has an interest in them. He can prevent their diversion to any purpose inconsistent with the bargain he has made, and if the corporation by its proper officers determines to divide them and does divide them, the individual shareholder's rights with respect to them do not then simply originate; they come to fruition in the final act, that has been aimed at from the beginning. The "dividend" he receives is an aliquot part of the fund divided; the fund itself is the source of the part that he receives, and if on analysis the fund is derived from various sources, some of which are within Australia and some outside Australia, he is, according to the provisions of the Act, liable or not liable to taxation in respect of it accordingly. The Act treats a dividend from profits arising in Australia as also arising in Australia. We say nothing as to the particular source - viz., (1) personal exertion or (2) property to which the dividend itself ought to be ascribed.

We have referred to these general considerations of law, not because we think they are at all necessary in the construction of the Act, for we do not think so - it is plain enough; we refer to them only to explain why, in our opinion, the arguments advanced on behalf of the appellant are insufficient to alter what, apart from them, is in our view the true meaning of the enactment read according to the ordinary and primary signification of its language. So reading it, we think it provides that every person deriving by way of income a share of profits made in Australia derives that income from an Australian source, whether he obtains that share as a partner in a firm, or as a member of an unincorporated association, or as a shareholder in an incorporated company. The substance is the same in each case, and we think the law is the same also.

The case will be remitted to the learned Judge, with the opinion of this Court that the sums referred to were rightly included in the assessment.

Case remitted to the Supreme Court with the opinion that the sums referred to were rightly included in the assessment. Costs of the special case to be costs in the appeal.

Solicitor for the respondent, Gordon H. Castle, Crown Solicitor for the Commonwealth.

[6] (1908) 2 K.B., 89, at pp. 97-98.

[12] [1915] UKHL 2; (1915) A.C., 1022, particularly at pp. 1037-1040, 1042.

[15] (1914) A.C., 11, at pp. 16-17.

[17] (1916) 2 A.C., 307, at p. 338.

[23] (1915) 1 Ch., 626, at pp. 635, 640.