Woodruff, J.
Moses Hold, the defendant in this case, was indicted, tried, and acquitted on the charge of having, as postmaster, embezzled money order funds of the United States on November 27th, 1907. This charge was not based upon any alleged conversion or conversions to his own use of specific items of such funds, but on the fact that in his remittance letter of November 29th, 1907, which purported to remit all money order funds on hand at the close of business November 27th, 1907, except the odd cents, he stated that the cash balance for November 27th, as shown by his cash book a't the close of business on that day was $1,114.81, whereas the cash book itself shows a balance of $2,090.11. Thus the charge of embezzlement in that case was based upon the prima facie showing of a shortage of $976.11.
Upon the same investigation which led to that indictment, the grand jury returned six other indictments against defendant, all based, like that upon which he was acquitted, on alleged shortages. The record evidence establishes these discrepancies for each case, from the books and remittance letters of defendant as postmaster, from the routine record of the post office business as shown by the tags and envelopes used for forwarding remittances from defendant’s offi.ee to Honolulu, together with the dating stamps impressed upon such tags and envelopes, and the official records of their arrival at the Honolulu post office, and from the sailing and arrival dates of Hawaiian Inter-Island mail-carrying steamers, during the period covered by the indictments.
The indictments based upon shortages, including the one already tried, are, as respects the dates of the alleged embezzlements or shortages and their amounts:
September 18, 1907, $ 875.25;
September 25, 1907, 943.35;
October 16, 1907, 1,129.53;
[464]*464October 24, 1907, 915.00;
1,865.00; November 8, 1907,
730.21; November 12, 1907,
976.11. November 27, 1907,
The case now before the court upon defendant’s special plea of former acquittal, and demurrer thereto by the prosecution, is different from the other cases of the above list in the one particular that the remittance letter dated November 2nd, 1907, does not allege a balance for August 31st, 1907, different from that shown by the cash book. Like all the other allied cases, however, this indictment does not enter into particulars, but the record evidence shows that as far as the amount charged as embezzled ($1,865) is concerned, it is based upon a previously existing shortage discovered on the date charged in the indictment November 8th. This was discovered and evidenced, as shown by the record made part of defendant’s special plea, by the declaration of defendant to Inspector Hare, made on November 8th in the course of official business, that he had remitted said $1,865 November 2nd, whereas the record shows, and is uncontradicted, that the remittance was not sent until several days after November 8th, thus establishing prima facie a shortage of $1,865 on November 8th.
Defendant’s plea of former acquittal must, if it has any value, be paramount to a declaration that Avhen he Avas tried for the alleged embezzlement of $976.11 on November 27th, 1907, he was likewise put in jeopardy as to any embezzlement involved in the alleged shortage of $1,865 on November 8th; and that the jury, under proper instructions from the court, could have, on the trial of the indictment particularizing November 27th, given a verdict based partly or entirely on the evidence adduced or a-ddueible concerning the indictment to which the special plea has been entered. If such a verdict could have been reached at the trial as a matter of law, the plea of former acquittal must be sustained and the defendant discharged. If, on the other hand, such a verdict could not have been reached laAvfully, defendant'was not placed in jeopardy concerning the [465]*465alleged embezzlement of November 8th, and bis special plea must be dismissed.
Let us follow briefly the history of the former trial. The prosecution was evidently confronted from the beginning of the trial with two difficulties; one real, namely, that the evidence adducible was not such as would show any specific conversion or conversions of government funds; the other merely apparent, namely, that to make out a case it would be necessary to show a system of embezzlements evidenced by a series of shortages, and false official statements. The showing of a system was not a necessity, since the establishment through the post office records kept by defendant himself, of a false statement made by him in an official paper November 29th concerning the balance on hand November 27th, combined with a failure to remit $976 of the funds on hand November 27th, established in itself a prima facie showing of embezzlement, which shifted the burden from the prosecution to the defense (charge of Judge Dole to the jury in previous trial). Evidence concerning several other shortages did not shift the burden any more completely but merely made it heavier.
Be that as it may, from the beginning of the trial the government struggled to place in evidence all the official books, letters, entries, post-marks, tags, envelopes, etc., which had to do with each and every shortage set forth in the above list. The defense contended against the admission of this evidence. Judge Dole, after long argument and consideration, decided (written opinion on file) that all evidence tending to show through manifest shortages or false official statements, that money order funds had been embezzled, was admissible. ILe naturally admitted also all official records prior to the date of the earliest indictment, for what they were worth either to the prosecution or the defense to* strengthen the circumstantial chain of evidence of the alleged system of embezzlement or to help the defense bear the burden of raising a reasonable doubt as to ■ whether the apparently intentional shortages and false [466]*466statements did as a matter of fact result from one or many items of embezzlement.
Briefly the main points of evidence relied on by the prosecution as shown by the record, were as follows:
1. That remittance letter of November 29th, established, by its false entry concerning the cash book balance of November 27th, a prima facie showing that, at some time on or before November 29th, $976.11 had been embezzled by defendant.
2. That although prior to the remittance letter of August 16th, 1907, no such letter contained, when compared with the cash book or as far as known, any false statement concerning the state of his money order funds or the amount actually remitted with such letter.
3. But the remittance letter of August 16th, 1907, and all subsequent letters up to and including that of November 29th, except only those purporting to remit the balance on hand as of August 31st, September 30th, and November 30th, contained, as compared with the cash book, false statements of the balance of money order funds on hand.
4. Although the remittance letter dated November 2nd, 1907, did not state falsely as compared with the cash book, the balance on hand October 31st, nevertheless a shortage as of November 8th or some time earlier of the entire $1,865 which should have been on hand October 31st, is shown by the following record evidence: Inspector Hare inspected defendant’s postoffice November 8th, 1907. He found an apparent shortage ($1,865) of money order funds. Defendant explained this by stating that he had remitted $1,865 November 2nd, 1907. Pursuant to his custom, Inspector Hare noted the money order funds shown to him by defendant as actually on hand (pp. 64, et seq). He found some cash, and checks aggregating $350.65. According to custom, he made a memorandum of the checks. Among them he noted two checks, for $83 and $10, dated respectively November 1st and 7th. These two checks, and one for $5 received by defendant on November 12th, without contradiction that they were in defendant’s post office on November 8th, or later, were shown, also without contradiction, to have been in the alleged remittance ($1,865) of November 2nd. There is further uncontradicted corroboration of the [467]*467prima facie conclusion to be drawn from the presence of these checks in the $1,865 remittance, based upon post-marks and official records (p. 74, ei seep), that the alleged remittance of November 2nd was actually despatched later than November 12th and therefore could not explain the shortage of $1,865, discovered by Inspector TIare on November 8th.
The defendant practically admitted the facts set forth above, but places evidence of record intended to raise a reasonable hypothesis contrary to the assumption of specific embezzlements :
1. Prior to August 1st, 1907, the date upon which the prosecution claimed the system of embezzlement began, every suspicious feature of the alleged system is in evidence from the exhibits of the prosecution, except only the one feature of stating officially and falsely a balance different from that shown by the cash book.
2. That during August and the next succeeding months the defendant was recovering from an illness. His weakness led him to delay making up his cash book until after the first of each month, whereby it was impossible to know, except for the remittance at the beginning of each month, during August, September, October and November, the actual cash book balance. Wherefore the hypothesis intimated is that for convenience the amount ready for remittance was temporarily treated in the remittance letter as a cash book balance, — a lie perhaps, but certainly not an embezzlement.
3. During his entire postmastership, defendant, to serve the public as well as possible, accepted at his own risk personal cheeks and government warrants, but understood that he must not remit small checks, especially those with odd cents, nor cash if he could possibly help it; also that he should not remit by his personal checks. Eor this reason defendant was constantly put to it to transform small checks, warrants and cash into reasonably large checks without odd cents, and for that purpose had his nephew, the assistant postmaster, Joshua D. Koki, carry a more or less perfunctory bank account with Bishop & Company. There was frequent and considerable delay in transforming cash, small checks and even government warrants into proper drafts. The bank was naturally reluct[468]*468ant to give credit before collections were made. The hypothesis intimated is that both before and after August 1st, including the magnified case of November 2nd, the suspicious feature of delay in remitting was not due to embezzlement of any government funds, but to the extreme efforts of defendant to carry on his post office business in the way in which the government officials desired, or perhaps to malfeasance by the assistant postmaster, Joshua D. Hold.
4. Defendant’s father died November 1st or 2nd, and the remittance letter of November 2nd was signed by Joshua D. Kolri, assistant postmaster. The hypothesis intimated is that on November 8th, when defendant told Inspector Hare that $1,865 had been remitted November 2nd, he had every reason to believe so' and as far as he is concerned at least had no knowledge of or part in that particular delay, being 'afflicted during early November by the double difficulty of his own physical weakness mentioned above, and the death of his father.
So much for the main facts of the case. It would require a volume to set forth all the details and subtleties of the evidence. Nearly every official transaction of defendant from August 11th, 1905, to December 13th, 1901, the date on which he ceased to hold his position, was placed before the jury against the repeated objections of defendant. The jury acquitted defendant. To reach that verdict required the unanimous votes of the twelve jurors, and since defendant did not deny any of the suspicious features brought forward by the prosecution’s evidence, it can only be concluded that the hypotheses intimated by the testimony of defendant, must have raised a reasonable doubt in the minds of each and every juror as to his guilt.
Rut was that verdict merely that they had a reasonable doubt as to whether the suspicious circumstances of the false statement in the letter of November 29th, proved an embezzlement at some time prior to that date, or did they, having before them many times more evidence concerning alleged shortages of other dates, particularly this one of November 8th, and concerning the system of business of defendant before and after [469]*469August 1st, conclude that they were reasonably doubtful as to whether defendant had at any time during incumbency of his office, embezzled money order funds?
Let us first turn to the charge to the jury to see whether the court thereby warranted the jurors in finding a verdict which covered the suspicious circumstances contained in all the indictments connected with this system of false statements.
In paragraph 16 the court says:
“ If you find from the evidence that the defendant so managed his operations as postmaster that the receipts of the office served to cover, for the time being, a distinct previous embezzlement, or embezzlments, such conduct would not be embezzlement.”
It is not necessary to consider whether the court was wrong in making this charge contrary to the views of the prosecution. The court charged in this way properly I believe, and thus removed the possibility that the jury could confine itself to such specifically proven and admitted instances as the use of the Akona check for $83, the two Vredenberg checks for $10 and $5 respectively, and the L. A. Andrews check for $40, as specific cases of embezzlement. The elimination of these items left the jury with no recourse but to convict or acquit on the shortage of November 27th, or, if the charge of the court warranted them in looking further, on some one or more of the other shortages.
In paragraphs 19 and 20 the court instructed the jury specifically that a belief on their part in the existence of a shortage on November 27th, must, in the absence of a reasonable doubt as to its meaning, lead to a verdict of guilty; and if the court had referred to the other shortages merely as evidence tending to show a system, we might conclude that, whether the court had gone as far in the charge as it should, the verdict of the jury under the instructions of the court, should have covered only the shortage of November 27th.
However, when we turn to paragraph 17 we find the court says:
[470]*470“ If you believe from these books of account and remittance letters [in evidence] that the defendant converted to his own use in any way whatsoever on or about November 27th, 1907, any of the money order funds of the United States referred to by books and letters and covered by the indictment, then you may return a verdict of guilty.”
Applying rules of legal construction to this charge, I am forced to the conclusion that paragraph 17 and paragraphs 19 and 20 are not merely repetition of the same idea. Paragraph 17, through the expressions " or about November 27th,” and "m any way whatsoever,” and “awy of the money order funds," leads irresistibly to the conclusion that the court was charging the jury that they might convict or acquit for embezzlements of which they were convinced other, than those which might have taken place November 27th. Thus, if the L. A. Andrews check for $40 paid to defendant for money orders "about November 27ih” namely, on November 29th, had been converted to private use by defendant on that date, the jury would have been warranted in basing a verdict upon that specific case of embezzlement. Is there any less reason to believe that the same would be true of the Akona and Yredenberg checks, even though tlie jury had been convinced that they were converted to private use at or about the time they were received by defendant November 1st, 7th, and 12th respectively? Is it not also true that the shortage of' $1,865 discovered by Inspector Hare on November 8th, the shortage of $730.21 on November 12th, as shown by the false statement in remittance letter of November 15th, and the shortage of November 27th, are all, in the light of the record evidence, inextricably interdependent upon each other?
Lot us look at the practical impossibility of determining from the uncontradicted record evidence, (a) the date upon which the alleged system of embezzlements began; (b) how many separate items of, embezzlement there were; (c) how many different shortages there were; (d) to what extent and when shortages caused by embezzlements were reduced by any [471]*471private funds of defendant; (e) the extent to which delayed remittances like those of August 9th, and September 13th, or false statements of balance, were caused not entirely or partly by specific embezzlements but by the uncontradicted effort of defendant to convert cash and small checks into desirable drafts.
Besides the above list of seven indictments, it would have been equally reasonable for the grand jury to return two others as far as false statements of balance are concerned, — one based on remittance letter of August 16th, which evidenced a shortage of $709 on August 14th, and the other August 23rd, evidencing a shortage of $571.48 on August 21st. The letter of August 16th is the first known false statement in defendant’s official records, and the prosecution assumes that the embezzlement must have therefore begun after July 31st. But turning to remittance letter of August 9th, which forwards the actual balance on hand July 31st, and to Government’s Exhibit No. 17, we find that the balance of $1,010.17 available for remittance July 31st could probably have been remitted on the Claudine, reaching Honolulu August 1st, and surely on the Kinau, JVIauna Loa, or Claudine, reaching Honolulu August 3rd, 6th and 8th respectively. Why this delay over three or perhaps four mail steamers ? Suspicions having once been aroused that a system of embezzlement began soon after July 31st, why not infer from this delay that it began before that date, and that the delay was caused by the necessity of waiting to obtain enough funds to cover the July 31st balance, either from money order receipts of the office or from the private resources of defendant? Again, granting that these well-warranted suspicions are correct, how, except by knowledge of specific items of embezzlement, could the prosecution find the initial shortage in the alleged system, or determine from time to time just what possible fluctuating shortages there might have been ? The false statement in remittance letter of August 16th is only additional evidence of some shortage at or before that date. Again turning to the remittance letter of September [472]*47213th, we find that the prosecution did not deem it necessary or wise to indict for a shortage based on the great delay in transmitting by that letter, although Government’s Exhibit 17 shows that the three mails received, in Honolulu September 5th, 7th and 12th were alloAved to go without the remittance. This delay, in view of the evidence that the system of embezzlement must have begun before August 16th, would surely have warranted an indictment for embezzlement on or about September 13th. The shortages evidenced by false statements in remittance letters, and by delays in remitting, are too uncertain in time and amount, and too much complicated Avith the explanatory eAudence of defendant to make it reasonably probable that any one of these shortages can be so extricated from the others, that a trial based upon one can be said to be reasonably separable from any other in the same system.
• If there were any embezzlements during November Ave can not, from the evidence, determine or guess exactly the amounts, the number, and the dates of any or all of them, to say nothing of attempting to find to what extent apparent shortages at any particular date, were affected by non-criminal efforts of defendant or his agent Joshua D. Koki to convert cash and small personal checks into larger drafts, or even by criminal misuse of the funds by Joshua D. Koki.
It is reasonable then to believe that “about November 27th” reaches back as a matter of laAV at least through the alleged period of systematic shortages following August 1st, 1907, which led to the indictment now before the court.
The only hesitation about reaching this conclusion from paragraph 17 of the charge, would be the italicised words, “and covered by the indictment.” But it would be absurd to assume that the indictment “covered” only embezzlements actually of November 27th. Time is not necessarily of the essence of a criminal charge. If the time “covered by the indictment” may be one day before or after NoArember 27th, it certainly can be two days, and I hold may cover reasonably all the [473]*473days from August 1st to December 13th, if that is tbe period of tbe alleged systematic embezzlements, — surely can cover to November 8th or earlier.
Therefore I conclude that the charge to the jury must have and should have led the jurors to take into account the voluminous evidence before them concerning the alleged shortage discovered by Inspector Hare on November 8th, 1907; and that the defendant was placed in jeopardy concerning that specific alleged shortage. The acquittal therefor at the previous trial reached to and included both in law and in fact the indictment now under consideration.
This conclusion could not have been reached if the discovery of a shortage, whether by inspection as in the case of November 8th or by a false official statement as in the case of November 27th, were in itself an embezzlement. Shortages are mere evidence of embezzlement, which can be rebutted as was shown by the verdict in the November 27th case. There has been a constantly progressive but thoroughly logical development of the judicial view of the crime of embezzlement, to recognize the difference between indictments based on shortages or like suspicious circumstances alone, or on specific items of conversion. At first it was natural for accused persons to argue that, since penal statutes must be construed strictly as against the government, no conviction for embezzlement could be had unless some specific act of converting trust funds to private use could be shown. This view, however, would frequently, and in egregious cases, defeat the intent of the statute. It is clear from the excellent treatise on construction of uenal statutes contained in United States v. Bitty (208 U. S. 393, 402-403), that whereas the rule of strict construction for penal statutes requires that any interpretation reached must be based upon words found in the statute itself rather than upon a judicial idea of what the legislators should have said; nevertheless, when the actual words of a statute can be interpreted reasonably to effectuate rather than to nullify the evident intent of Congress, they should be construed in that way. Sec. 4046 R. S. makes [474]*474it embezzlement for any postmaster to convert to his own use, in any way, any money order funds of the United States. This and other statutes concerning embezzlement say nothing about proving specific acts of embezzlement. The words of this statute, therefore, simply require the prosecution to convince the jury beyond a reasonable doubt that there must have been a conversion of trust funds.
Thus the prosecution has, in all cases of embezzlement, its choice between indictments based upon specific acts of conversion, or based upon shortages or like suspicious circumstances. These shortages or suspicious circumstances, if more than one has been discovered for the same defendant, may be of two kinds, either those which are inextricably interdependent, or those which can be conclusively separated from each other. Thus a collector for several firms might be short in his accounts with two or more of his employers, and might thereby be subject to indictment and trial for more than one shortage, whether those shortages became apparent in one or many books of account, and in one day or from time to time. If, however, the shortages resulted from the same trust relationship and belonged evidently to the same system, and one or more of the specific items of embezzlement which caused the shortages could not be proved, the prosecution would seem to be in the predicament of basing its success on one trial, and the greater the number of shortages or false statements, the greater the burden on the defendant to raise a reasonable doubt.
Having chosen from necessity or otherwise to indict upon shortages, the prosecution cannot evade the constitutional safeguard against double jeopardy by the subterfuge of trying the accused more than one time upon various indictments, which in substance are not charges of embezzlement of the amount of the shortages, but merely charges that there must have been an embezzlement or embezzlements at some time because of the existence of shortages.
In other words the prosecution may legitimately treat embezzlement, when evidenced by shortages, delays in transmit[475]*475ting, etc., as a quasi “continuing offense” (State v. Reinhart, 38 Pac. 822, 821), a “single embezzlement” (Brown v. State, 18 Ohio 496), or as unidentifiable items of embezzlement to be proven only by the “aggregate result” (Ker v. People, 110 Ill. 646), through the proof of shortages, delays, etc., which can not be reasonably explained.
To compel this choice is fair to the prosecution because it does not allow the unfaithful agent to escape punishment merely because he is ingenious enough to hide the specific items which constitute the “continuing offense1,” the “single embezzlement,” or the “aggregate result.” If the prosecution can discover the specific items of embezzlement, it can, without resort to the indictment for provable shortages, have as many indictments or counts as the specific items of embezzlement discovered. If, on the other hand, the law compels, the agent to assume the burden of proof against the prima facie case made out by the establishment of a shortage, it is only fair to him that all interrelated and interdependent shortages up1 to the date of the indictment, should be tried once and for all in the same case, as is the proper practice in other continuing offenses such as unlawful cohabitation, etc.
If this conclusion is correct the reluctance of the court to take what appears to be a further step than that already reached in progressive cases like State v. Reinhart, Brown v. State, and Ker v. People, is counterbalanced by satisfaction that the decision is in direct line with the evident intention of the constitutional provision against repeated jeopardy, and protects this defendant not only from the harassment and expense of a second trial in which it is apparent the evidence must be practically the same as in the previous case, which occupied about five weeks, but further from the possibility of six further trials upon existing indictments of the same defendant upon the same evidence provided he should be successively acquitted, to say nothing of the further possible indictments on such suspicious circumstances as the false statements of August 16th and 23rd, and the delays of August 9th and September 13th. The”object [476]*476of the constitutional provision against repeated jeopardy was certainly aimed at the possibility of cases like this. Unless defendant has resources beyond that which appears in the record evidence in the last trial, he would very soon be forced to the position of trusting his defense to assigned counsel, of languishing in jail for want of resources to obtain bail, and of trusting to the charity of the government for the various costs which so unfortunately follow all trials.
The demurrer is overruled.