This is a bill in equity against the Waltham Watch Company and its directors, brought by a minority stockholder of the defendant company in behalf of himself, the company and all other stockholders. The bill seeks a decree establishing the obligation of the company and its directors under its agreement of association to declare and pay dividends to the class A common stockholders and to declare and pay or accumulate dividends on the six per cent preferred stock annually whenever the operations of the defendant company result in annual net earnings, and establishing the obligation of the directors to declare and pay dividends on the common stock, class A, and to declare and pay or accumulate dividends upon the six per cent preferred stock for the.years 1939, 1940 and 1941. The pertinent parts of the agreement of association are annexed to and made a part of the bill. The case was heard upon statements of agreed facts, which the parties agreed are all the facts. The judge entered a decree dismissing the bill, from which the plaintiff appealed.
A summary of the facts follows: The plaintiff is the owner of nine hundred shares of the six per cent preferred stock of the defendant company, and has owned these shares continuously since the reorganization to be mentioned hereinafter. The individual defendants are the directors of the *399company and were the directors during the years 1939, 1940 and 1941. The company was organized and incorporated in Massachusetts in 1923 to succeed to the watch-making business of a predecessor company of the same name. The old company met serious financial difficulties at the end of the last war, having accumulated a large inventory, much of which was obsolete and overvalued. It had little cash and was forced to resort to bank loans and public offerings of five-year coupon notes to keep going, meanwhile continuing to operate at a loss. The old company exhausted its borrowing power and could obtain no further extensions on its bank loans, and despite all efforts continued to lose money until the bank creditors approached Kidder, Peabody & Co. to undertake a financial reorganization. The assets of the old company were sold to the new defendant company in 1923; the new company assumed the debts of the old company and procured cash to liquidate the $7,200,000 of bank loans and publicly held coupon notes by the sale for $5,300,000 of the bonds, debentures, and a portion of the six per cent preferred stock of the new company, by the issue of $1,700,000 of prior preference stock for subscription at par by the old preferred stockholders, and by the issue of common stock, class A, for $250,000. (For full details of the reorganization, see Abbot v. Waltham Watch Co. 260 Mass. 81.)
Since the date of the reorganization, the financial position of the new company has steadily improved. Earnings have been sufficient to enable the company to retire its debentures without refinancing, and to purchase and retire substantial amounts of its mortgage bonds and stocks, so that on December 31, 1941, there were outstanding the following securities: $1,130,000 of mortgage bonds due June, 1943; prior preference stock of a par value of $381,230; six per cent preferred stock of a par value of $3,384,285; twenty-five thousand shares of no par value common stock, class A; and forty-one thousand eight hundred sixty-eight and three fourths shares of common stock, class B, no par value.
The pertinent provisions of the agreement of association and articles of organization of the company follow: “Sec*400ond: Holders of prior preference stock shall be entitled to receive, when declared by the Directors, cumulative dividends at the rate of 7% per annum and no more .... No dividends shall be paid on any other class of stock until all accrued dividends and the current quarterly dividend on the prior preference stock shall have been paid, or a sufficient sum from which to make payment when due shall have been set aside for that purpose. Third: Holders of preferred stock shall be entitled to receive, when declared by the Directors, dividends at the rate of 6% per annum and no more, payable before any dividends are paid on the common stock in any year, except as hereinafter provided in the case of common stock, Class A. Such dividends shall be non-cumulative except as hereinafter provided. Fourth: Holders of common stock, Class A, will be entitled to receive each year in dividends an amount equal in the aggregate to one-fifth of the net earnings of the Company for the preceding year, to be determined as follows: At the end of each calendar year the Directors shall ascertain the annual net earnings for the preceding year remaining after the payment of all interest charges, accrued and unpaid dividends on the prior preference stock and after providing during the first five years for the annual sinking fund, to retire the debentures dated February 15, 1923, and shall forthwith declare and pay to the holders of the common stock, Class A, a dividend equal in amount to one-fifth of such net annual earnings, provided the Company’s capital will not be impaired by such payment. If the Company’s capital will be so impaired, then the Directors shall forthwith declare and pay a dividend to the holders of the common stock, Class A, of so much of the one-fifth of such net annual earnings as may be paid without impairing the Company’s capital. The rights of the holders of common stock, Class A, to the dividends above provided for shall be noncumulative. If a dividend as above stated is paid on the common stock, Class A, in any year, there shall either be paid in the same year on the preferred stock a dividend equal in amount to four times the dividend paid on the common stock, Class A, but not exceeding an amount equal *401to 6% per annum on the preferred stock outstanding, or if such dividend is not paid during the same year, it shall be accumulated, and no dividends shall be paid on the common stock, Class B in any year unless such accumulated dividend on the preferred stock is paid as well as the dividends on the preferred stock for such year. . . . Ninth: All the stock, both preferred and common, shall have equal voting powers, each share entitling the holder to one vote.”
During the years 1939, 1940 and 1941, the company had an earned surplus and, after the payment of taxes and all charges, had earnings in excess of the amount necessary to pay all accrued dividends on the prior preference stock. The figures for 1939 and 1940 are shown in the balance sheets and earnings statements for those years attached to the record. No figures were available for 1941 when the statement of agreed facts was prepared. All cumulative dividends on the prior preference stock for the years 1939, 1940 and 1941 have been paid. It is conceded that after such payments net earnings were sufficient in those years for the declaration and the payment of a dividend in some amount on the common stock, class A, and on the six per cent preferred stock; that during the years in question there has been no impairment of capital in the sense of “contributed capital” or “capital represented by outstanding capital stock”; and that out of the earnings of each of the three years dividends in some amount could have been declared and paid on the class A common stock and on the six per cent preferred stock without causing an impairment of capital in that sense. At the close of each of the years in question the directors have ascertained the amount of earnings remaining after provision for accrued dividends on the prior preference stock, but have determined to declare no dividends on the class A common or the six per cent preferred stock.
The parties are agreed that the sole issue is whether, under the agreement of association and articles of organization set out above, it is within the discretion of the directors to declare no dividends on the common stock, class A, and on the six per cent preferred stock in years in which there *402are an earned surplus and net earnings of the prior year available therefor, or whether, as the plaintiff contends, the agreement of association makes a declaration mandatory in such circumstances. If the declaration of dividends be held discretionary, the plaintiff does not charge abuse of discretion.
It is well settled that a stockholder’s rights between himself and the corporation and the other stockholders are contractual, and that the terms of the contract are to be found in the agreement of association and the provisions of applicable statutes. Page v. Whittenton Manuf. Co. 211 Mass. 424, 427. Lee v. Fisk, 222 Mass. 418, 420. Thomas v. Laconia Car Co. 251 Mass. 529, 533. Wabash Railway v. Barclay, 280 U. S. 197, 202. Collins v. Portland Electric Power Co. 12 Fed. (2d) 671, 673. Therefore it is to the agreement of association, particularly clause fourth thereof, that the court must look for a determination of the plaintiff’s right to receive and the defendants’ duty to declare dividends.
The cases in this Commonwealth as well as those of other jurisdictions reveal, in general, a reluctance on the part of the courts to construe provisions relative to the declaration of dividends in such a way as to hold that it is mandatory upon the directors in any circumstances to declare dividends. The courts prefer not to interfere thus with the sound financial management of the corporation by its directors, but declare as a general rule that the declaration of dividends rests within the sound discretion of the directors, refusing to interfere with their determination unless a plain abuse of discretion is made to appear. Field v. Lamson & Goodnow Manuf. Co. 162 Mass. 388, 393, et seq. Fernald v. Frank Ridlon Co. 246 Mass. 64, 71-72, and cases there cited. Thomas v. Laconia Car Co. 251 Mass. 529, 535. Morse v. Boston & Maine Railroad, 263 Mass. 308, 311-312. Wabash Railway v. Barclay, 280 U. S. 197, 202-204. Collins v. Portland Electric Power Co. 12 Fed. (2d) 671, 672. See also 18 C. J. S. 1106, where, citing authorities, it is said: “The mere fact that a corporation has a large amount of surplus or net profits does not entitle the stockholders to the pay*403ment of dividends. When a corporation has net or surplus profits, unless some restraint is imposed by statute, charter, by-laws, contract, or otherwise, whether a dividend shall be declared and, if declared, its amount, rest in the sound discretion of the directors.”
This principle is applicable with equal force to the rights of holders of preferred shares, whether cumulative or noncumulative, to receive dividends. The cases just cited were all cases of preferred stock. In Morse v. Boston & Maine Railroad, Wabash Railway v. Barclay, and Collins v. Portland Electric Power Co., just cited, the shares in question were all noncumulative preferred shares. See also New York, Lake Erie & Western Railroad v. Nickals, 119 U. S. 296, 307. As was said in the Collins case, 12 Fed. (2d) 671, 673: “Preferred stock does not represent a debt nor a pledge of profits in favor of the holders thereof in preference to others.”
Field v. Lamson & Goodnow Manuf. Co. 162 Mass. 388, affords an excellent example of the disinclination of our court toward a construction that the payment of dividends is mandatory. In that case, under a statute that permitted the corporation to guarantee dividends on its preferred shares, the corporation issued, certificates bearing the following language: “The holder of the.stock represented by this certificate is entitled to dividends thereon annually out of net profits, in preference and priority to the holders of any stock of said corporation except the preferred stock issued under said act, to the amount of six per cent .... The holder of the stock represented by this certificate is entitled to dividends upon it at six per cent for each year from the time, of its issue, cumulatively, before any dividends shall be paid upon any stock of said corporation except the preferred stock issued under said act, which dividends are guaranteed by said company” (pages 391-392). There it was held that to construe the guaranty contained in the stock certificate in the usual sense attributed to a guaranty would be contrary to the provision of the statute that dividends were to be paid only out of net profits, and the court did not choose to construe the lan*404guage as a guaranty that a dividend would be paid if there were net profits. The court saw in the language “nothing ... to take the case out of the general rule, that, in the first instance, the decision of the question whether there shall or shall not be dividends lies with the company or its agents” (page 394). This, we think, is as far as this court has gone to avoid a construction requiring the declaration of dividends.
It is true that some cases may be found where a court has disagreed with the views expressed above. Thus, in Belfast & Moosehead Lake Railroad v. Belfast, 77 Maine, 445, the by-law there involved provided: “Dividends on the preferred stock shall first be made semi-annually from the net earnings of said road, not exceeding six per centum per annum, after which dividend, if there shall remain a surplus, a dividend shall be made upon the nonpreferred stock up to a like per cent per annum ...” and in that case it was held (page 452) that “The language of the bylaw is really the language of the general law. It promises dividends whenever there are net earnings from which to make them.” This was followed, as correctly stating the law as to noncumulative preferred shares, in Hazeltine v. Belfast & Moosehead Lake Railroad, 79 Maine, 411, a case involving the same issue of stock and the same by-law as that in issue in Belfast & Moosehead Lake Railroad v. Belfast. See also Spear v. Rockland-Rockport Lime Co. 113 Maine, 285. But Belfast & Moosehead Lake Railroad v. Belfast, 77 Maine, 445, with other similar cases cited, was expressly disapproved by this court in Morse v. Boston & Maine Railroad, 263 Mass. 308, where the court, after referring to the Belfast case and other like cases, said at page 312: “We do not think, however, that these cases should be followed; on principle as well as on authority we believe the better rule to be that the declaration of dividends on the noncumulative preferred stock is within the sound discretion of the directors, taking into account the finances of the company, its obligations, its condition and all the matters which were for the consideration of the directors.”
However, as has been indicated at the beginning of this *405discussion, it. is agreed that the rights of a shareholder rest on contract; and if the contract, as expressed in the charter, the by-laws, or a statute, plainly requires the construction that in certain circumstances dividends must be paid, the court will, as it must, adopt that construction and establish the obligation of the directors to declare a dividend, for there is no positive rule of law forbidding such a provision.
The only Massachusetts case which we have been able to discover where the language of the contract, contained in a by-law, was found to impose an obligation on the corporation to pay dividends is Lydia E. Pinkham Medicine Co. v. Gove, 303 Mass. 1. The text of the by-law is set forth in 303 Mass. at pages 11-12. In Cratty v. Peoria Law Library Association, 219 Ill. 516, the court held that the following by-law made the payment of dividends mandatory. The by-law there in issue provided: “The association guarantees to every stockholder a dividend of eight per cent per annum on the amount of paid-in stock held by him or them from the date of such payment, which dividend shall be a charge against the association, and the production of the certificate of stock or receipt of payment shall entitle the holder to the amount of such dividend. Such dividend shall be due and payable on the first day of January in each year.” The court there held that “It certainly cannot be said that the parties to this contract intended that it should not be performed if the corporation preferred to expend its net income in increasing the capital or assets by adding to the library and rendering it more useful and valuable” (page 523).
In Burk v. Ottawa Gas & Electric Co. 87 Kans. 6, 15, the court held that the following by-law made mandatory the payment of dividends in accordance with its terms. The by-law provided: “The preferred stock shall carry a six per cent per annum preferred, noncumulative dividend, payable semiannually on the first days of July and January of each year after January 1st, 1906, out of the net profits of the preceding fiscal year, and a pro tanto dividend if such dividend fall short of 6 per cent.” In that case the court said, *406“It is against public policy to construe the terms upon which stock is issued so as to require dividends to be paid to preferred creditors out of the invested capital, as that might result in the destruction of the corporation. (Lockhart v. Van Alstyne, 31 Mich. 76, 18 Am. Rep. 156.) But no such objection attaches to a construction that makes mandatory the payment of dividends out of the net profits” (pages 15-16). The court was aided in reaching that construction by another by-law which showed that, when dividends were meant to rest in the discretion of the directors, the draftsman knew how to employ appropriate language. That by-law read thus: “Upon the common stock only such dividends and at such times shall be paid out of the company’s net profits as the board of directors may in their judgment deem it advisable to declare.”
In the case at bar the defendants argue that the apparently mandatory language of clause fourth of the agreement of association must nevertheless be held to leave the matter of dividends out of net profits on the common stock, class A, and on the six per cent preferred stock within their judgment. They strongly stress the predilection of the courts for such a construction and rely upon the apparent inconsistency that will exist between that clause and clause second relating to dividends on the prior preference stock if the plaintiff’s interpretation be the one adopted. Clause second reads: “Holders of prior preference stock shall be entitled to receive, when declared by the Directors, cumulative dividends at the rate of 7% per annum and no more, commencing to accrue on the 12th day of March, 1923, and payable July 1, 1923, and quarterly thereafter on the first days of January, April, July and October in each year. No dividends shall be paid on any other class of stock until all accrued dividends and the current quarterly dividend on the prior preference stock shall have been paid, or a sufficient sum from which to make payment when due shall have been set aside for that purpose.”
It is argued by the defendants and conceded by the. plaintiff that the declaration and the payment of dividends on the prior preference stock are expressly made subject to the *407directors’ discretion. Yet both accrued and current dividends on that stock must be paid before any dividend ihay be declared on any other class of stock. From this the defendants maintain that the declaration or the payment of dividends on the common stock, class A, and on the six per cent preferred stock must be discretionary with the directors. In our opinion, however, the language of clause fourth of the agreement of association is so direct and express that we cannot escape the conclusion that it requires the directors to declare dividends out of net earnings of the preceding year in the amount and under the terms stipulated on the common stock, class A, and on the six per cent preferred stock. It is difficult to conceive of words more direct and expressive of command than the following language used in this clause: “At the end of each calendar year the Directors shall ascertain the annual net earnings for the preceding year . . . and shall forthwith declare and pay to the holders of the common stock, Class A, a dividend . . . provided the Company’s capital will not be impaired by such payment. If the Company’s capital will be so impaired, then the Directors shall forthwith declare and pay a dividend to the holders of the common stock, Class A, of so much . . . .”
“Shall forthwith” suggests immediacy and necessity; it is expressive of the imperative; it negatives any suggestion of discretion or determination of policy by the directors. The entire clause suggests irresistibly that the policy of the directors with regard to net earnings has been fixed and declared for them beyond their power to alter. In short, it is apparent that the incorporators determined to dedicate a certain amount of such net earnings as the company might have to the payment of dividends on the class A common and on the six per cent preferred shares of the company.
Further support for this conclusion is found, as it was found in Burk v. Ottawa Gas & Electric Co. 87 Kans. 6, in the fact that when it was intended to confer discretion on the directors in the matter of dividends the draftsman of the agreement knew how to, and did, employ suitable language. Thus, when it comes to declaring dividends on the common *408stock, class B, clause fifth provides that “Any earnings at any time applicable to the payment of dividends after the payment of the dividends in respect of the prior preference stock, preferred stock and common stock, Class A, herein-before specified, may be applied to the payment of dividends on the common stock, Class B . . . .” One has but to contrast the phrase “shall forthwith declare and pay” with the words “may be applied to the payment” to ascertain the obvious difference of intent expressed in the two clauses of the agreement.
The history of the origin of the defendant company tends to indicate that the plaintiff’s construction of the agreement of association is the true one. The common stock, class A, and the six per cent preferred shares were sold to the public as the issue of a company succeeding to the business of a failing concern, in an effort to raise new capital for the business. It seems most likely that new investors in a business whose future was certainly speculative could not be induced to risk their money in the new company except upon the assurance that if operations should result in net earnings they, the investors, would definitely receive a share of those earnings. In view of the risky nature of the investment it seems most probable that purchasers of the shares put their entire hopes in a share of any net earnings that might result; that the possibility of permanently augmenting their wealth was sufficiently remote at the time for them to be unwilling to permit the directors to accumulate a large surplus or plow all the earnings back into the business while they, the investors, were foregoing any return on their investment during years when such return was possible. In short, the following language employed in Burk v. Ottawa Gas & Electric Co. 87 Kans. 6, seems applicable to the case at bar. It is there said: “The directors of the corporation owed a positive duty to pay a dividend to the preferred stockholders whenever in any year there were net profits available. The funds that might be used for that purpose could not rightfully be expended for extensions merely for the benefit of the business, nor could they be withheld to meet the expenses of the next year. Inasmuch as the only possible source of *409profit to the preferred stockholder from his investment is the distribution of earnings in the year in which they accrue, he has a right to insist that an accounting shall be taken annually, and that the surplus of one year, available for a dividend, shall not be carried over to meet a possible deficiency of the next” (page 16).
It is not difficult to reconcile this construction of clause fourth with the provisions of clause second dealing with the payment of dividends on the prior preference stock. Reading the agreement of association as a whole, the interpretation suggested by the plaintiff seems the one most likely to have been truly intended by the incorporators. It brings all the provisions into harmony with each other, and gives effect to each clause without altering the meaning most naturally to be placed upon the words employed in each. And it is neither strained nor artificial. The plaintiff’s interpretation is, briefly: In years when there have been no net earnings the directors may in their discretion pay dividends on the prior preference stock out of accumulated surplus, or they may pass them; likewise in years when earnings are meager and insufficient to pay a dividend on the prior preference stock and leave anything more. Similar discretion in like circumstances is granted to the directors by clause third with reference to the payment of noncumulative dividends on the six per cent preferred stock. But in years when the operations of the company result in net earnings sufficient, after the payment of certain prior charges, to pay a dividend on the prior preference stock with something left over, then a dividend on the prior preference stock must be declared and paid or a sufficient sum therefor set aside, and, further, a dividend of the stipulated portion of the remainder declared and paid to the holders of the common stock, class A, and a cumulative dividend of a stated portion of such remainder declared in favor of the six per cent preferred shares.
Such an interpretation does not make it impossible for the directors ever to accumulate a surplus out of earnings to guard against future contingencies, or to reinvest a part of the earnings in the business. In those years when net *410earnings remain after paying the dividends required on the prior preference, the six per cent preferred and the common, class A, stocks, whether or in what amount dividends shall be paid to the holders of the class B common stock is a matter for the determination of the directors. They may declare no dividends out of that balance or declare a dividend equalling less than the entire balance. The absolute rights of the holders of the common stock, class A, and of the holders of the six per cent preferred stock are limited to a stated share of net earnings in years when net earnings are available. There is nothing in the agreement of association that confers on any class of stockholders of the company a right to dividends out of earned surplus. Nor ought it to be held, in the absence of the clearest expression of intent, that the incorporators had so little regard for future contingencies that might arise that they drew an agreement making it impossible in any circumstances for the company to amass a surplus to provide for unforeseen emergencies or to tide the company over lean years.
One point remains to be disposed of. Clause fourth of the agreement of association requires one fifth of the net earnings of the preceding year to be declared and paid as a dividend to holders of the common stock, class A, “provided the Company’s capital will not be impaired by such payment.” If the company’s capital will be impaired, then a dividend is to be paid “of so much of the one-fifth of such net annual earnings as may be paid without impairing the Company’s capital.” The defendants argue that the word “capital” as here employed means “working capital.” The plaintiff insists that the word means “contributed capital.”
An examination of the cases in which the word “capital” has been employed and defined indicates that the courts are not all agreed on the meaning of that word. The definitions run all the way from capital stock liability, that is, par value of the stock, to the total of all the assets, tangible and intangible, owned by the corporation, with various intermediate degrees. The term, however, seems never to have been employed in the sense of “working capital” when capital impairment is the subject under discussion. When *411“working capital,” that is, cash and other quick assets, is intended, that term is employed. To hold that “capital” here means “working capital” would completely nullify the plain purpose of clause fourth, for, as the plaintiff points out, no payment of dividends or' any disbursement of any kind could ever be made without to that extent impairing the working capital of the company. If the prohibition of the statute against the payment of dividends out of capital were to be construed as meaning capital in the sense of working capital, rather than in the sense of contributed capital, no dividend could ever be paid since all of the assets of the company are capital in the broad sense until they are separated and divided by the company. Minot v. Paine, 99 Mass. 101, 106. But the accumulation of surplus profits does not increase the capital unless new shares are issued as against profits so accumulated. Gray v. Hemenway, 212 Mass. 239, 242-243. Old Colony Trust Co. v. Jameson, 256 Mass. 179, 182, 183, and cases cited. Gibbons v. Mahon, 136 U. S. 549.
We are of opinion that capital in some sense other than working capital was in the minds of the incorporators when they drew up the articles of association, for the problem of capital impairment by declaration of dividends is one of ancient and common consideration. The prohibition was announced as a common law rule as long ago as 1824 in Wood v. Dummer, 3 Mason, 308. It was a common provision of special acts granting corporate charters, one of the first being found thirty years before the decision in Wood v. Dummer, in the charter of the Insurance Company of North America in 1794. (See Kehl on Corporate Dividends, c. 1, § 6.) There has been legislation on the subject in many States, particularly in recent years but going all the way back to Laws of New York, 1825. There is no reason to believe that, when language appropriate to this very common policy of dividend regulation is employed in a situation where one would expect to find the subject under consideration, the language was used in the most unusual sense of “working capital”.
What, then, is the meaning of the word “capital” as here *412used? Without discussing the various definitions of that word by courts of other jurisdictions, we may rely on that word as defined in the Massachusetts taxation cases as giving at once the simplest and, we believe, the most realistic meaning. The word has been passed upon in many cases in deciding what is capital and what is income for the purpose of an income tax levied on dividends out of profits but not out of capital by G. L. (Ter. Ed.) c. 62, § 1, subsection (g). This has been adjudicated many times, but it is not necessary to search beyond the two recent cases of Commissioner of Corporations & Taxation v. Filoon, 310 Mass. 374, and Flint v. Commissioner of Corporations & Taxation, 312 Mass. 204, where the problem was thoroughly considered and definitely disposed of. The word “capital” is thus defined in the Filoon case: “the word ‘capital’ may have different meanings when used in different connections. . . . It appears, however, from the Putnam case [Tax Commissioner v. Putnam, 227 Mass. 522] and the cases following it, and from the general purpose of G. L. (Ter. Ed.) c. 62, § 1, subsection (g), that the word ‘capital’ as used therein means property invested in the corporation by the stockholders, and is not limited in its meaning to ‘capital’ in the narrow sense, sometimes described as ‘legal capital,’ that is, property sufficient to balance the capital stock liability” (page 382). See also Flint v. Commissioner of Corporations & Taxation, 312 Mass. 204. “Capital,” for the purposes of the tax statute, is distinguished in these two cases from net profits, accumulated profits or earned surplus, out of which dividends are taxable as income. Accumulated profits are defined as being true accretions to the corporate wealth as distinguished from “capital” under the statute, which refers to the sum contributed by the shareholders to legal capital. Commissioner of Corporations & Taxation v. Filoon, 310 Mass. 374, 385-386. Flint v. Commissioner of Corporations & Taxation, 312 Mass. 204, 206.
It is conceded by the parties that the word means either “working capital” or “contributed capital.” We have already expressed the opinion that it does not mean the former, and we think that it must be taken to mean “con*413tributed capital,” that is, the amount invested by the shareholders — “capital” as the word is employed in G. L. (Ter. Ed.) c. 62, § 1, subsection (g). And this seems the most natural construction because it is a fair inference that the parties were thinking not in dry terms of accounting concepts but of protecting the actual investment of the shareholders when the agreement of association was framed. It follows, therefore, that it should be held that dividends on the common stock, class A, and on the six per cent preferred stock are required to be declared out of net earnings of any preceding year if the company’s contributed capital will not be impaired thereby.
It might be argued that, since the agreement of association in question speaks of “annual net earnings for the preceding year” as the fund out of which dividends are to be declared, it could not have been intended to use the words “net earnings,” as defined in the Filoon and Flint cases, as a true increase in the corporate wealth; and that the meaning of the words as here employed is only an increase in the corporate wealth over the preceding year, regardless of the fact that in prior years the capital of the company has been impaired, so that the net earnings of the preceding year do not represent a true accretion to the corporate wealth over the sum contributed by the shareholders. But the answer to this is that the agreement requires that the annual net earnings be distributed as dividends “provided the Company’s capital will not be impaired by such payment.” If such net earnings of the preceding year are to be taken to mean merely an increase in the corporate wealth over the preceding year, regardless of any prior losses and capital impairments not made good, the words “by such payment” would be a nullity, for on such a construction the distribution of the net earnings could never work a capital impairment. It seems obvious that the incorporators intended that capital impairments be made good before the net earnings of any year could be distributed as dividends to shareholders, and that any payment of dividends out of net earnings before capital impairments were made good would ipso facto be a capital impairment. Further support is given to this *414construction by the provision that if a distribution of the net earnings will impair the company’s capital, then the directors shall distribute so much of the net earnings among the shareholders as will not impair the company’s capital. If the language is to have any meaning it must mean that, if any net earnings remain after making up previous capital impairments, they, and they only, may be distributed as dividends, and that the distribution of any greater amount of the net earnings would impair the company’s capital because it would not make good prior impairments.
This construction of the language in question is not novel. In 18 C. J. S. 1101, it is stated as a rule of general application, with citation of authorities, that “dividends may not be paid from profits of the current year where there remains a deficit from losses sustained in previous years. The actual value of the assets with which the corporation began business is to be deducted and not the nominal or par share capital, and the business of a corporation is to be viewed as a unit in determining whether or not there are net profits.” The New York stock corporation law, § 58, as amended by Laws of New York, 1939, c. 364, expressly provides that “No stock corporation shall declare or pay any dividend which shall impair its capital, nor while its capital is impaired . . . .”
The case of Branch v. Kaiser, 291 Penn. St. 543, is of particular interest. A Pennsylvania statute now repealed as to business corporations (Act of May 23, 1913, Pub. Laws, 336,) provided that dividends “shall in no case exceed the amount of the net profits actually acquired by the company, so that the capital stock shall never be impaired thereby.” In that case the court said (page 550): “It is argued by counsel for the respondents that the impairment of the capital occurred prior to the declaration of the dividends in question, and all that the directors did was to fail to reduce an impairment that existed. This argument overlooks the fact, however, that by declaring the dividends each year out of current profits, the directors illegally diverted funds which properly were applicable to a reduction of the capital deficit. The earned profits, therefore, reduced the impair-, *415ment, and the declaration of the dividend again impaired the capital to the extent of the dividends declared.” We think that this is the true construction of the language of clause fourth of the agreement in the case at bar. This conclusion is in harmony with the decisions in the Filoon and Flint cases, supra, which, as has been previously pointed out, define “capital” for the purposes of G. L. (Ter. Ed.) c. 62, § 1, subsection (g), in the same way in which, we hold, capital must be defined in this case. In Flint v. Commissioner of Corporations & Taxation, 312 Mass. 204, the court, quoting from the Filoon case, said: “If at any date losses exceeded prior ‘accumulated profits’ the deficiency in such profits — that is, the excess of losses over prior ‘ accumulated profits’ — must be made good before the corporation can be said to have ‘accumulated profits’” (page 206). And in the same case, the court said that “the shareholders receiving dividends, before the deficiency in ‘accumulated profits’ had been made good, would be receiving, in the guise of ‘accumulated profits’ subject to income taxation, property that had been invested by shareholders in the corporation and not earnings of the corporation representing an increase in the wealth of the corporation ...” (page 209). See also Spiegel v. Beacon Participations, Inc. 297 Mass. 398, 427.
While clause fourth of the agreement of association of the company relative to dividends on the six per cent preferred stock, does not provide specifically that they may be paid only if its capital will not be impaired thereby, we think that that restriction is implicit as a matter of law throughout the clause, otherwise the settled principles forbidding the payment of dividends which would impair the contributed capital of a company as defined in our earlier decisions to which we have referred would be violated. Since in the instant case upon the agreed facts it is unquestioned that the declaration of dividends for the years 1939, 1940, and 1941 would not impair the contributed capital of the company, we conclude that a declaration thereof in some amount must be made by the defendant directors of the company.
The decree entered in the court below is reversed and instead a final decree must be entered establishing the obli*416gation of the defendant directors of the company to declare and pay dividends on the class A common stock and to declare and pay, or accumulate, dividends on the six per cent preferred stock for the years 1939, 1940 and 1941, as provided in clause fourth of the agreement of association to the extent that this can be done without impairing the contributed capital of the company. The plaintiff is to have costs.
Ordered accordingly.