Brewer, J.
This action is upon county warrants. The circumstances which surround it make it one of peculiar importance. For 10 years, and from the admission of the state in 1876, many counties, whose tax levy was insufficient to meet current, expenses, had been issuing warrants, which had accumulated to, as counsel says, at least §1,000,000. No question seems to have been made during these years as to the validity of such action; but the question being thereafter presented to the supremo court of the state, it construed section 6, art. 11, of the constitution of the state in such a manner as to invalidate the bulk of these warrants. The plaintiff, being a non-resident, comes into the federal court, and invokes Its judgment as an independent tribunal on the question thus determined by the state supreme court.
No more delicate or responsible duty is ever cast upon the federal court than when it is asked to determine, not what the state supreme court has deckled, but whether its decision shall be followed. While the federal courts are in a certain sense independent tribunals, yet they sit within the state to construe and enforce its laws. Whenever a construction has been placed upon a state statute or constitution by her supreme court, that determines for both state and federal courts all questions and rights arising after such construction; but, when the rights or claims of right arose prior to such construction, then the duty rests upon the federal courts of independent examination and determination. The rule controlling in such cases is fully and clearly stated in the recent case of Burgess v. Seligman, 107 U. S. 33, 2 Sup. Ct. Rep. 10, as follows:
“Since the ordinary administration of the law is carried on by the state courts, it necessarily happens that by the course of their decisions certain rules are established which become rules of property and action in the state, and [846]*846have all the effect of law, and which it would be wrong to disturb. This is especially true with regard to the law of real estate, and the construction of state constitutions and statutes. Such established rules are always regarded by the federal courts, no less than by the state courts themselves, as authoritative declarations of what the law is. But where the law has not been thus settled, it is the right and duty of the federal courts to exercise their own judgment, as-they also always do in reference to the doctrine of commercial law and general jurisprudence. So when contracts and transactions have been entered into, and rights have accrued thereon under a particular state of the decisions, or when there has been no decision, of the state tribunals, the federal courts properly claim the right to adopt their own interpretation of the law applicable to the case, although a different interpretation may be adopted by the state courts after such rights have accrued. But even in such cases, for the sake of harmony, and to avoid confusion, the federal courts will lean towards an agreement of views with the state courts if the question seems to them balanced with doubt. Acting on these principles, founded as they are on comity and good sense, the courts of the United States, without sacrificing their own dignity as independent tribunals, endeavor to avoid, and in most cases do avoid, any unseemly conflict with the well-considered decisions of the state courts. As, however, the very object of giving to the national courts jurisdiction to administer the laws of the states in controversies between citizens of different states w'as to institute independent tribunals, which it might be supposed would be unaffected by local prejudices and sectional views, it would be a dereliction of their duty not to exercise an independent judgment in cases not foreclosed by previous adjudication.”
See, also, Pana v. Bowler, 107 U. S. 540, 2 Sup. Ct. Rep. 704; Green Co. v. Conners, 109 U. S. 104, 3 Sup. Ct. Rep. 69; Carroll Co. v. Smith, 111 U. S. 556, 4 Sup. Ct. Rep. 539; Anderson v. Santa Anna, 116 U. S. 362, 6 Sup. Ct. Rep. 413; Bolles v. Brimfield, 120 U. S. 759, 7 Sup. Ct. Rep. 736. Accepting that as the rule to control this court, I pass to a statement and consideration of the immediate question. The section of the constitution referred to is as follows:
“No county shall contract any debt by loan in any form except for the purpose of erecting necessary public buildings, making or repairing public roads and bridges;- and such indebtedness contracted in any one year shall not exceed the rates upon the taxable property in such county following, to-wit: Counties in which the assessed valuation of taxable property shall exceed $5,000,000, $1.50 on each $1,000 thereof; counties in which such valuation shall be less than $5,000,000, $3 on each $1,000 thereof. And the aggregate amount of indebtedness of any county lor all purposes, exclusive of debts contracted before the adoption of this constitution, shall not at any time exceed twice the amount above herein limited, unless when, in manner provided by law, the question of incurring such debt shall, at a general election, be submitted to such of the qualified electors of such county as in the year last preceding such election shall have paid a tax upon property assessed to them in such county, and a majority of those voting thereon shall vote in favor of incurring the debt; but the bonds, if any be issued therefor, shall not run less than ten years, and the aggregate amount of debts so contracted shall not at any time exceed twice the rate upon the valuation last herein mentioned: provided, that this section shall not apply to counties having a valuation of less than $1,000,000.”
Tbe decisions of the supreme court of Colorado referred to are People v. May, 9 Colo. 80, 404, 414, 10 Pac. Rep. 641, 12 Pac. Rep. 838, and [847]*84715 Pac. Rep. 36.1 In the first of these cases the question was whether this section in all its sentences referred exclusively to debts contracted by loan, and it was held that it did not; that there were two independent declarations in it; and that the third sentence, commencing, “And the aggregate amount of indebtedness of any county for all purposes,” was not limited to debts by loan, but applied generally to all. In the second case the question was whether the limitation upon county indebtedness imposed by this section included debts incurred by operation of law as well as those arising from express contracts, and it was held that it did.
Returning now to the first question, it was and is earnestly insisted by the learned counsel for plaintiff that the scope and intent of the entire section is indicated by its opening lines, which expressly name debts by loan; that, as a general rule, each separate section is to be construed as carrying one idea through all its sentences, and that, even without special words of reference in the later sentences, they are to be construed as based upon the first idea or controlling fact; that not only is such the ordinary rule of construction, hut also to establish a different one for this section would result in glaring absurdities, hindering the administration of county affairs in such a manner and to such an extent as shows that such was not tire intent of the framers of the constitution or the people who adopted it.
Again, it is insisted that although, prior to the framing of this constitution, seven states had incorporated into their constitutions limitations upon county indebtedness, a comparison of the condition of those states with this, in territory, population, and county organization, shows the absurdity of the meager powers given to counties in this state under such a construction of this section. And, again, great reliance is placed upon contemporary construction, inasmuch as for 10 years many counties have transacted business and issued warrants as though no such limitation as is now claimed existed; inasmuch as several suits were brought in different courts, both state and federal, upon county indebtedness, in which the defense of this limitation existed, but was not presented; inasmuch, also, as several acts of the legislature applied to facts as they existed, and as it must be presumed wore known to exist, conflict with the limitation imposed by this construction.
It will be observed that the language of the third sentence is general. It says: “That the aggregate amount of indebtedness of any county for all purposes, exclusive of debts contracted before the adoption of this constitution, shall not at any time exceed,” etc. blow, in order to make it mean what counsel claim it means, the word “such,” or some word of similar moaning, must bo implied prior to the word “indebtedness” or “purposes;” and counsel insist that such word should be implied, and that the meaning is correctly expressed, when the language is “the aggregate amount of such indebtedness,” or “for all such purposes.” That a word may be supplied or implied when necessary to carry out the ob[848]*848Vious intent of tbe sentence is, of course, conceded; but can the intent be presumed to justify the supply of the omission? It is, of course, always a question of construction, and the thought and intent of the framer is the thing to be determined. I have read the opinion of the supreme; court on this matter, and studied it with care. The solid tread of the argument of the learned judge who wrote that opinion is to my mind irresistible. It compels conviction. Starling with the conceded and elementary proposition that, in determining the meaning of any instrument, whether agreement, statute, or constitution, we look first to the words used, and seek their natural signification in the order of grammatical arrangement in which they are found, he shows that the sentences are separate and complete, the propositions in them independent declarations, and that there is no grammatical necessity of implying or supplying any word to perfect either. Hence the supplying or implying of a word must find some other reason than the mere grammatical arrangement for its justification.
Again, he calls attention to what, in my judgment, is most significant, — the proceedings of the convention that framed this constitution. If, in construing any sentence, it becomes a question whether any word should be implied which is not found, in order to fully express the true meaning, and we can know that when the sentence was being prepared the question of whether that very word should be inserted was presented, discussed, and it was finally determined to reject it, we should be driven almost irresistibly to the conviction that the meaning -which that omitted word would disclose was not the meaning intended by the makers. Now, it appears from the debates in the convention that the question oí introducing this word “such” into the sentence was presented and dis. cussed; that two or three times it was voted to introduce it, and as often voted to strike it out; and that finally'it was left out, and the proviso at the close of the section added. I do not see how demonstration could be made more perfect as to the intent of the framers of this section. Cooley, Const. Lim. 66. That which was their intent as shown by the grammatical arrangement of the language, and by their discussion, must, in a matter of this kind, be presumed to have been the intent of the people in adopting this constitution. The learned judge also notices other matters, such as the language of an address prejuired by the convention, and submitted with the constitution to the people; but I do not care to pursue this matter further, or notice the various reasons urged by him in support of the conclusion. It is enough to say thatT think his reasoning unanswerable.
I pass-now to the second question: Does the limitation upon county indebtedness imposed by this section include debts incurred by operation of law as well as those arising from express contracts? This question may be really separated into two: First, in determining when tin limit is reached, what is to be included; and, second, what effect has the limit when reached upon the powers and liabilities of the county?
In regard to the first, it will be noticed that county indebtedness may arise in one of three ways: It may spring from the voluntary contracts [849]*849of tbe county authorities. Second, it may be cast upon the county by tbe action of the legislature in requiring tbe payment by it of certain foes and salaries; and in respect to debts of this class the county as such exercises no choice, has no volition. They are compulsory, as distinguished from contractual, obligations. Further, it may arise from some tortious act of the county authorities. Such a tort creates a liability against the county, which, ripening into a judgment, becomes a debt. Now, in determining when the limit of county indebtedness has been reached, it is obviously immaterial under said section how any particular portion of the indebtedness arose or whence it sprang. It is enough that it is a debt. The language of the section is, “the aggregate amount of the indebtedness of any county for all purposes shall not exceed,” etc. Every dollar of the indebtedness might have sprung from tortious acts of county officials which, prosecuted by the injured parties in actions ex delicto, have ripened into judgments, and thus become debts, and still tbe limit be reached. Indeed, the process of determination is a mere matter of mathematical calculation, — the adding up of the valid debts of the county.
Secondly. Supposing the limit has been reached. What effect has that on the powers and liabilities of the county? That it puts an end to its contractual powers is clear. So the supreme court hold. Such was the obvious purpose of the framers of the section. They were familiar with the experiences of other states, the territorial laws in force, and the ordinary ways of county business everywhere. They knew that a largo discretion was given to county commissioners in the matter of creating obligations against tho county. They knew that it was usual to give to them power to build and keep in repair county buildings, to lay out, open, and improve county roads, build bridges, and otherwise subject tbe county to large expenditures. With this in view, the obvious purpose of the statute was to limit their powers in this direction, and to say that beyond a certain sum no county official should be authorized to contract a debt binding against the county. On the other hand, it is equally clear that this section does not enable a county to defeat its liability for a tort, or prevent such liability from being merged into a judgment, and thus becoming a debt. It is no defense to an action for a wrong that, if judgment be recovered therefor, it will carry the county indebtedness beyond the constitutional limit. This is conceded by tbe supreme court in this opinion. It says: “Involuntary liability arising ex delicto is a subject that is not contemplated by the provision.” In this tho court only follows prior decisions in other states. In Bartle v. City of Des Moines, 38 Iowa, 414, the same doctrine was affirmed. So, also, in the case of Bloomington v. Perdue, 99 Ill. 329, the court in that case in a singlo sentence disposing of the question, as though a contrary doctrine were too absurd to require discussion. See, also, Chicago v. Sexton, 115 Ill. 230, 2 N. E. Rep. 263, in which the court uses this language:
“There is nothing new in thus holding a municipality responsible for the want of fidelity of those who act for it. Suits of that kind are of daily occurrence. The liability thus imposed is not within the constitutional and statutory limitations in regard to the creation of indebtedness.”
[850]*850So it has been held in Iowa that where warrants were issued in excess of the constitutional limit, and judgments rendered thereon, the defense of invalidity not having been pleaded, and that through the fraud of the supervisors of the county, and thereafter bonds issued in payment of such judgments, the bonds were valid in the hands of a bona fide holder. Railroad Co. v. Osceola Co., 45 Iowa, 168, 52 Iowa, 26, and 2 N. W. Rep. 593.
The other class of debts springs from neither the voluntary nor the tor-tious acts of county officials. The county has neither voice nor opportunity in the matter. They are imposed by the legislature, and are generally such as affect the state at large as well as the county.
It is well here to stop a moment, and consider what a county is. In one aspect, it is an independent corporation, having peculiar private interests and concerns. The management of those interests and concerns is, as a general rule, confided to the county officials; and the debts incurred in the management of those private affairs are created by the voluntary contracts of those officials. In another aspect, the county is but a mere subdivision of the state, and only determines locally the administration of those affairs which affect the people of the state as a whole. Take the administration of justice in the courts, the matter of elections, the preservation of the public peace, and matters of a kindred nature. They are not the purely private concerns of the county. They affect, vitally and largely the interests of the state as a whole. It is common elsewhere, it was and is the case here, that the cost of these public services is cast largely upon the county. Not upon the count}*- as an independent corporation, and solely interested in and benefited by such services, but as a portion of the state, and as such portion thus contributing to the general welfare. In the creation of debts for these services the county is not consulted. It has no voice in saying when they shall be incurred, or to what extent. I know the line of demarcation is not preserved with absolute uniformity, but the general character of the difference between contractual and compulsory obligations is as I have stated. This is a matter of common knowledge, and must have been within the contemplation of the framers of the constitution. Was it their intent to relieve the count}* of liability for these compulsory obligations when in any manner the general limit of indebtedness had been reached? See what that would imply: The possibility that- county commissioners, by extravagance, might largely impair, if not practically defeat, the administration of justice, the preservation of the peace, and even the holding of elections. For public service, without expectation of pay, is seldom done, or, if done, only poorly. Will a constable serve process; will a sheriff, at personal risk, preserve the public peace; will a county attorney prosecute with vigor and interest; will juror or witness attend, giving up private interests for public good, — with the knowledge that these, their services, are gratuitous, and will receive no compensation?
Will it be said the legislator can guard against these dangers by an increase of the tax levy? Who shall say how great will be the extravagance of the county officials? or, even if that could be restricted by law, [851]*851who can foretell how much of crime will happen during any coming year? I am advised by counsel that the suppression of a single riot in Loadville cost the county more than the entire proceeds of that year’s levy. The;.e things cannot be foreseen. Can it be that the framers of the constitution intended that compensation for such services should depend upon the question whether a limit of county indebtedness had been reached? Tt must be borne in mind that the amount of such expenditures can never he foreseen. In one year they may aggregate a great amount, and in the next be comparatively trifling.
These considerations of a general nature impress me forcibly with i ho conviction that it was not the intent of the framers of this section to permit a county to escapo from liability on account of such compulsory obligations by the fact that its general limit of indebtedness has been reached. This conviction is strengthened by tins provision in lire section: that tiro indebtedness “shall not at anytime exceed twice the amount above herein limited, unless when in manner provided by law the question of incurring such debt shall at a general election be submitted,” etc. Now, what a county proposes voluntarily to do, what amount of debt it intends voluntarily to create, may be known and submitted to a vote; but who can foresee what amount of crime may bo committed, what criminal cases may have to bo prosecuted, what expenses must be incurred in such prosecution, or in preserving the public peace, not yet apparently threatened? These are matters which, as they cannot be foreseen, cannot well be provided for by submission to popular vote. “Such debt,” is the language; obviously a debt that is contemplated, and not one that may be imagined. This question did not arise in this state for the iirst time. 1 n 1885 the supreme court of Missouri considered and decided it. In Book v. Earl, 87 Mo. 246, that court held that a comity could not contract a debt for any purpose in excess of its revenue for the current year without the special assent of the voters, and that, when it was desired to create a debt in excess of such revenue for the improvement of the court-house, the question must first be submitted to the qualified voters. But at the same term, in Potter v. Douglas Co., Id. 239, iho same court held that the constitutional limitation as to indebtedness had no application to a debt incurred by a county for the keeping and transporting of its prisoners by the sheriff or jailor of another county. The judge who wrote the opinion in that case thus states the reasons therefor:
“ After carefully considering the subject, I am not of opinion that the constitutional prohibition should be ruled to apply in instances like liie present. For this conclusion these are my reasons: I do not regard section 12, supra, as applying here, because the effect of such construction would be destructive of the peace and good order in every county embraced within the provisions of section 6090, aforesaid; for it would be an impossibility to submit to a vote of the people of the county concerned, the question of an unascertained and unasccrtaiuable indebtedness to be incurred in the future, as the exigencies of the ease might demand. Who could foretell how many criminals would be arrested in the course of the ensuing year? If this could not be done, is it not glaringly obvious that no question as to the amount of the indebtedness could possibly be submitted to the people for the sanction of their suffrages? [852]*852The maxim, lex non aogit ad impossibilia, may appositely be invoked in the present case; a maxim equally invoeable whether the law be statutory or organic. But another reason occurs why that section cannot apply in the case at bar. The inhibition of the constitution, it will be observed, is leveled against a county becoming indebted, i. e., through the ordinary channel, the action of the county court, the financial agent of the county. But here the indebtedness was not so incurred. It was created entirely independent of any action of the county court; created by the sheriff of the county pursuant to the command of section 6090, supra. The law itself gave license to the incurring of such debt. It was incurred by operation of law, and the fact that the county would ultimately have the debt to pay cuts no figure in this discussion.”
It is true, by the constitution of Missouri there is a limit on the amount of taxation, as well as in the matter of indebtedness, but that does not militate against the force of the argument presented by that court. My conclusion, therefore, is that it is no defense to an action upon county warrants issued in payment of these compulsory obligations that the general limit of county indebtedness has been reached. I believe that it is agreed between counsel that the warrants sued on in this action are of this nature, and the judgment must be entered, therefore, in favor of the plaintiff..
I regret that in the conclusions that have been reached I am not fully in accord with the supreme court of the state. > I regret this, not alone because I regard it generally the duty of the federal court to follow the supreme court of the state in its interpretation of the state statutes and constitution, but also because of the high respect I entertain for the individual members of that court, who honor the office which they occupy, and whose duties they so admirably discharge; but over against that is the strong voice of appeal rising from the equities of these various warrant holders. This is not the case, of a series of bonds issued to a railroad corporation whose existence and'work are, notwithstanding it is called a quasi public servant, we all know, prompted by personal and selfish interests, but it is the case of warrants issued to many individuals, witnesses, jurors, constables, sheriffs, and the like, called upon by the laws of the state to render some public service in the administration of justice, the punishment of crime, the preservation of the public peace, the conduct of elections, — matters in which the state as a whole is interested and which conduce to the general welfare, — and who have obeyed the mandates of their state in the undoubting faith that the promised compensation for those services would be paid. Honesty, justice, and all the potential obligations of a solemn promise demand that they be paid; and if I have lent a too-willing ear to the voice of such appeal, and have permitted such equities to blind my eye to a clear vision of the literal mandate of the constitution, it is consolation to know that the amount in controversy is so great that a review of these proceedings can be had in the supreme court of the United States, by which high and ultimate tribunal any errors and mistakes can be corrected.