Olivee, Chief Judge:
The merchandise herein consists of glass perfume bottles, assessed for duty under paragraph 218 (e), Tariff Act of 1930, as bottles produced otherwise than by automatic machine, at 75 per centum ad valorem (in some cases, the assessment was made under the same paragraph, as modified by the trade agreement with Czechoslovakia, T. D. 49458, at 37% per centum ad valorem). They are claimed properly dutiable under the same paragraph at 25 per centum ad valorem as bottles “produced by automatic machine.” Paragraph 218 (e), so far as is pertinent, reads as follows:
Bottles and jars, wholly or in chief value of glass, of the character used or designed to be used as containers of perfume, talcum powder, toilet water, or other toilet preparations; * * * all the foregoing produced by automatic machine, 25 per centum ad valorem; otherwise produced, 75 per centum ad va-lorem. * * *
The same issue was before us in Jos. Riedel Glass Works, Inc. v. United States, 12 Cust. Ct. 173, C. D. 849. In that case, the bottles before the court were produced by an automatic machine. The contention of the Government was that, as the raw material was gathered by manual labor, rather than supplied automatically, the machine was not automatic; that to be automatic, it must be fully automatic, not semiautomatic or partly automatic. Upon the record, we found the bottles to have been produced by automatic machine and so held.
An appeal was duly taken and the decision of this court was affirmed (United States v. Jos. Riedel Glass Works, Inc., 32 C. C. P. A. (Customs) 201, C. A. D. 307). The appellate court there said:
Appellant concedes that so far as the imported merchandise is concerned, the actual forming of the molten glass into bottles was an automatic process, but contends that the bottles were not produced by automatic machine because the services of a gatherer were required in feeding the molten glass into the machine.
The court cited with approval the following definition from Webster’s New International Dictionary, Second Edition, 1936:
automatic machine. A machine of machine tool which, after once being set, operates automatically except for applying the power, lubrication, supplying material, and shutting off the power. [Italics quoted.]
[200]*200The court added:
* * * it is clear from the record that the machine which produced the bottles in this case operated automatically, after once being set, except for supplying the raw material — molten glass. * * * Once the Taw material is placed into the machine, all other operations are automatic, and the next thing that emerges is a completed bottle. [Italics supplied.]
Subsequently, a new case was tried (Griffon Importing Co. v. United States, 20 Cust. Ct. 179, C. D. 1106). The only evidence submitted by the plaintiff therein was a deposition from a representative of the English manufacturer. Unlike the machine in the Riedel case, supra, which automatically produced the bottles, the “machine” in the Grifón case, supra, required, in addition to the gatherer of the molten glass, two operators who, by hand, controlled, opened, and closed the molds, introduced compressed air into the molds, removed the partly formed bottles at intermediate stages from one mold to another, and finally, manually removed the completely formed bottles to a conveyor belt for transfer to the annealing oven (lehr). Each operation was independent of the succeeding operation. No two operations were interdependent or mechanically hooked together, and each operation called for the judgment, experience, and manual operation of a skilled operator. Because this record showed a complete absence of the automatic features of the machine in the Riedel case, supra, and a substitution of full manual control of every single operation, this court held that bottles so manufactured were not produced by automatic machine but were otherwise produced and were therefore subject to the higher rate of duty as classified. This decision was by the majority of the division, one judge dissenting. On appeal, the decision of this court was reversed (Griffon Importing Co. v. United States, 36 C. C. P. A. (Customs) 121, C. A. D. 408). The majority opinion (O’Connell, J.) was concurred in by two associate judges, the late Judge Hatfield filing a separate concurring opinion. A lengthy dissenting opinion, filed by Judge Jackson, was concurred in by Judge Johnson.
In the present case, the perfume bottles were produced by the same process as the bottles in the Grifón case, supra, which was by use of the so-called Schiller-type machines. Plaintiffs incorporated the record in the Griffon case, supra, and offered no further testimony. The Government called five additional witnesses and introduced in evidence a number of exhibits. Among the exhibits, are samples of a bottle in various stages of its production, models of the molds, and photographs of the so-called Schiller-type machine. There is also before us a motion picture in color, showing, in operation, both the so-called Schiller-type machine and also a fully automatic bottle-making machine. This motion-picture - film was projected on a [201]*201screen for the benefit of tbe court and counsel. As in tbe Riedel and Griffon cases, supra, plaintiffs claimed tbe bottles properly dutiable at 25 per centum ad valorem as having been “produced by automatic machine.” As hereinbefore stated, tbe only evidence in support of tbe protests in tbe Griffon case, supra, and in tbe present case, is a deposition by one John Arthur Maier connected with tbe English manufacturer. In answer to interrogatory No. 4, the witness described tbe operation of the so-called machine, used to make tbe bottles there and here before us, as follows:
a. Molten glass is transferred, by a gatherer using a solid gathering iron, from the furnace to the parison mould to which are fitted the neekring moulds.
b. No. 1 Operator cuts off sufficient glass which, by means of vacuum created by continuously working pumps to which the machine is connected, is automatically sucked into the neck moulds. This operation forms the aperture, the brim and the external shape of the neck of the bottle.
c. The quantity of glass is largely determined by the size or capacity of the parison mould and also partly by the point at which the operator cuts the glass from the gathering iron.
d. Compressed air, supplied from power driven compressors, is released through a control valve to blow the internal shape of the neck and the parison form. The parison form is determined by the shape of the parison mould.
e. The neekring moulds in their spring loaded holders, supporting the partly formed bottle, are transferred (by No. 1 Operator) to the blow or finishing mould. No. 2 Operator lowers the blow head thereby releasing the valve which controls the compressed air supply. The bottle is then automatically blown to the shape of the finishing mould. The finished bottle is removed from the mould by No. 2 Operator who compresses the spring of the neck mould holders, automatically ejecting the bottle on to the table of the machine. From this point the bottle is carried to the annealing lehr.
Tbis Griffon case, supra, was clearly distinguishable from tbe Riedel case, supra, where the entire operation of tbe machine was automatic after tbe gob of molten glass bad been gathered and cut off by manual labor. In tbe Griffon case, supra, on the other band, not one single operation of the entire cycle of producing the bottle was automatic.
In tbe Griffon decision, supra, we said that from plaintiff’s testimony, tbe deposition, it appeared that tbe “bottles were produced by machine,” and that tbe compressed air was “supplied by tbe machine.” Tbis is not factually correct. It appears that no bottles were produced by machine and tbe compressed ab was supplied, not by tbe “machine,” but by “continuously working pumps to which tbe machine is connected.”
Our appellate court in tbe prevailing opinion in tbe Griffon case, supra, said (page 125):
* * * That the word “automatic” has different meanings under different circumstances * * *.
[202]*202That is undoubtedly so, but we know of no instance where a machine which performs no single operation automatically, that is by itself, has been known as, or held to be, an automatic machine. The court went exhaustively into the congressional hearings in connection with paragraph 218 (e) of the 1930 act. It will serve no purpose to repeat that discussion here.
The description of the operation of the “machine” in the Griffon case,' supra (and in the present case), as contained in the deposition above referred to, is in many respects incomplete and misleading. The present record with its many exhibits and additional oral testimony clarifies many points which may not heretofore have been clear. While the operation of the so-called English Schiller-type machine and those used in this country vary in some unimportant features, the following, gathered from the present record, can be said to fairly describe the operation of this “machine.” To avoid confusion, we refer to the 3-man team as the gatherer, operator No. 1, and operator No. 2, even though in some of the testimony they are referred to as operators Nos. 1, 2, and 3.
The molten glass is gathered on the end of a solid rod (punty) by one member of the 3-man team, called the gatherer. The molten mass so gathered is removed from the furnace by the gatherer who holds it over the first (parison) mold.
Sitting before the parison mold, operator No. 1 inserts a neckring mold into the bottom of the parison mold previously opened by him. He then lowers a funnel over the top of the mold to guide the glass into the mold. Using shears, he then cuts off sufficient glass to make, the bottle being manufactured. This gob of glass falls into the mold. The operator then raises a plunger in the neckring, lowers the blow-down head down over the mold, and admits compressed air which forces the glass down around the plunger. He then lowers a baffle plate down on the top of the mold. He then releases compressed air into the mold through the neckring and finally returns all movable parts to their original positions and opens the mold. Each and every one of the multiple operations above described is accomplished by operator No. 1 by means of various levers all operated by hand. No single movement or operation is performed by the “machine" itself. When the above-described activities of operator No. 1 are completed, and he has opened the parison mold, he removes the neckring mold, with the partly formed bottle in it, and transfers it manually to the finishing mold operated by his co-worker (operator No. 2) seated beside him.
This second operator (No. 2) meanwhile has opened one of the two finishing molds before him to receive the partly formed bottle transferred to him by operator No. 1. The finishing mold operator (No. 2) [203]*203has two finishing molds which he uses alternately for speed of operation and to permit each bottle to cool sufficiently to keep its shape when removed. When the partly finished bottle in the neckring mold is placed in the finishing mold, the operator closes the mold and locks it in position. He then moves the blowhead down on top of the mold and introduces compressed air into the mold. This blows the bottle to its finished shape. He then shuts off the air, raises the blowhead, opens the mold, and removes the completely formed bottle from the mold by means of the handles on the neckring mold. By compressing the handles of the neckring mold, the bottle is ejected and placed on a conveyor belt by means of which it is carried to the annealing oven (lehr). All these operations are performed by operator No. 2 by means of hand levers or foot pedals.
. The above-detailed description of the operation of what is called the “Schiller-type machine” presents a very different picture from that which was before the court in the Griffon case, sufra.
In the light of the augmented record now before us, we are of opinion that it has now been established, beyond any question, that the so-called machines used to produce the bottles now before us were neither “fully automatic” nor “semiautomatic” and cannot, by the most flexible application of the doctrine of congressional intent, be held to be automatic machines.
We have, throughout this opinion, referred to what the witnesses called machines as “so-called machines” or “machines,”, because we entertain grave doubts as to whether this collection of molds, levers, and compressed ah- tubing constitutes a machine. We have in the present case as additional evidence certain physical exhibits which were not before the court in the Griffon case, sufra. The still photographs of the “machine” (defendant's exhibits 15 to 22, inclusive), show a steel table mounted on wheels, making it mobile, and while the purpose is not set forth in the record, we may assume that this was so that the entire apparatus could be moved from one furnace to another, wherever the desired supply of raw material (molten glass) was available. On this steel table, and affixed thereto, are three bottle molds. - One is the parison mold which forms the neck of the bottle. • The other two are finishing molds in which the bottles are blown to their finished size and shape. So far as this record is concerned, and this is confirmed by a careful inspection of both the still photographs and the motion-picture film (defendant’s exhibit 14), there is no connection oj any kind between the various molds. So far as coordination or integration of motion is concerned, the several units could be located in separate rooms. The action of one mold has no connection with, or effect on, either of the other molds. There are no belts', gears, drive shafts, motors, or any of the other conventional [204]*204earmarks of a machine here present. The only power apparently employed is compressed air used to blow the glass in the molds. ' Not one single operation is automatic. Not one single operation is performed by the machine. It may be assumed that the group of three molds on the table were placed in juxtaposition solely for convenience of the operators who were working with hot material and must pass the partly finished product from one to the other. These units were placed close to each other but were not in any manner connected. In “A History of Mechanical Inventions,” by Usher, (p. 66) there appears the following:
Every machine will be found to consist of a train of pieces connected together in various ways, so that if one be made to move, they all receive a motion, the relation of which to that of the first is governed by the nature of the connection.
The so-called machine before us does not respond to this definition» as the “pieces” are not “connected together” in any way and the movement of one part does pot in any manner affect the operation of any other part. If the operators at the parison and finishing molds did not manually open and close their molds and by manual operation control the compressed air to do the blowing, no bottle would ever come into being. The skill and dexterity exercised by the operators of these molds, as evidenced by the motion picture before us (defendant’s exhibit 14), is of a very high order and is far removed from the action of an automatic machine which produces bottles without the intervention of any human control. The fully automatic machine will produce about 2,000 dozen 1-ounce bottles per day. This so-called 2-man machine will produce possibly 240 to 250 dozen in the same period (R. 54). In Simon, Buhler & Baumann (Inc) v. United States, 8 Ct. Cust. Appls. 273, T. D. 37537, a machine was defined as—
* * * a mechanical contrivance for utilizing, applying, or modifying energy or force or for the transmission of motion * * *.
This definition has been quoted with approval many times since it first appeared in the above decision of our appellate court in 1918. In view of some of the articles which, from time to time, have been held to be dutiable as machines, we hesitate to hold this group of unconnected molds not to be a machine. We do, however, point out that while power, to wit, compressed air, is used in connection with these molds, there is no evidence that the compressed air operates the molds. The molds are manipulated solely by the action of the operators. In a concurring opinion filed in our appellate court decision in the Griffon case, supra, the late Judge Hatfield said (page 130):
* * * It has been suggested that the machines involved in the instant case are not machines at all. I assume it is meant by that suggestion that, as [205]*205the glass bottles formed by the involved machines are manually removed to an annealing oven for the annealing process which, it clearly appears from the record is a necessary process in the production of glass bottles, the involved machines do not produce the involved glass bottles. [Italics supplied.]
The above is not the reason why we suggest that this apparatus may not be a machine. We are in grave doubt on this point because no "glass bottles” were “formed by the involved machines.”
As to the automatic feature, Judge Hatfield quoted a lengthy definition from Machinery’s Encyclopedia, Yol. 1, pages 178-179. We quote only part of this definition:
* * * When a machine is capable of automatically producing duplicate parts repeatedly, it is universally referred to as automatic, whereas, if it simply performs a complete cycle of machining operations, but requires the attention of an operator each time a part is finished, it may be considered automatic by some, and semi-automatic by others. In some cases, a machine of the latter class is termed “automatic,” while one that is capable of continuous operations is known as “fully automatic.” [Italics quoted.]
This so-called Schiller-type machine does not come within the above definition of an automatic or semiautomatic machine. It performs no machine operations at any stage. Its operators do not give it attention after it completes its appointed task. They perform the task, using the molds and compressed air as tools in the process.
To sustain plaintiffs’ protests, we must find that the Congress did not mean what it said in enacting paragraph 218 (e) and must substitute for the language used, other and different language because of certain statements made in the hearings prior to the enactment of the statute. As we said in the Griffon case, supra, (page 182):
* * * It is not within our province to make the laws. We may not substitute our opinion for the written statute. We must apply the language which was actually used by the Congress and have no right to give any meaning to such language other than that conveyed by the words, terms or expressions in which the legislative will was expressed. (Edgar Allen Steel Co. (Inc.) et al. v. United States, 16 Ct. Cust. Appls. 26, page 29, T. D. 42715.)
The language used by Congress is “ * * * bottles * * * produced by automatic machine, 25 per centum ad valorem; otherwise produced, 75 per centum ad valorem.” Whether or not the apparatus used to make the bottles before us can be said to be a machine, it cannot, by any stretch of the imagination, be held to be an “ automatic” machine. It is neither fully automatic nor semiautomatic and does not operate automatically in any single particular. In its decision in the Riedel case, supra, our appellate court said (page 204):
* * * Suffice to say * * * the intent of Congress as demonstrated by the legislative history of the statute was to protect American glass blowers against the competition of lower-paid glass blowers of other nations * * *.
This the Congress has done. Hand-blown bottles are subject to the higher rate of duty. The Congress, as we see it, has gone one step [206]*206farther and has also extended protection to the highly, skilled workers who produce bottles by the method here before us. The only part of the industry which does not need protection is that which produces bottles by automatic machine.
The following appears in the record of the hearings before the Committee on Ways and Means, Tariff Readjustment, 1929 (Yol. II, p. 1366):
Mr. Ramseyer. You agree witli the witness who preceded you, that there is no need for an increased tariff on machine-made bottles?
Mr. Gayner. In so far as I understand it.
Mr. Ramseyer. Do you agree with him that there should be some increase on handmade bottles?
Mr. Gayner. Absolutely; because with the machine-made bottles the labor element does not enter into it. The glass flows automatically into the machine. It is handled entirely by the machine and conveyed to the annealing lehr, and not touched by human hands until it is ready to be sorted and packed.
It thus appears that the committee had from at least one witness a description of the type of machine on which no additional protection was needed. It is the automatic machine where “the labor element does not enter into it”; where the bottle is “handled entirely by the machine” and conveyed to the annealing lehr, and “not touched by human hands” until it is ready to be sorted and packed.
It is our considered opinion that Congress was fully cognizant of the effect of its language when it specifically provided for the lower rate of duty on bottles “produced by automatic machine.” Had Congress intended the higher rate to apply only to hand-blown bottles, it could have said so. Had it intended to have the lower rate apply to all bottles made in any part by the use of compressed air as against mouth-blown, it could have said so. The fact is that after lengthy hearings, the Congress used the language now before us. There may be some doubt as to whether it intended to limit the lower rate to the product of a fully automatic machine or to externl it to cover also bottles produced on semiautomatic machines. We may not, however, apply the term “automatic” to something which has no single automatic feature connected with it.
The excerpt from the deposition in the Griffon case, supra, herein-before set forth, describing the process of manufacturing the bottles before us, sets forth in part that the glass “b. * * * is automatically sucked into the neck moulds * * *” and “e. * * * The bottle is then automatically blown to the shape of the finishing mould * * *” and, further, “The finished bottle is removed from the mould by No. 2 Operator who compresses the spring of the neck mould holders, automatically ejecting the bottle on to the table of the machine.” [Italics supplied.] The foregoing is the only testimony adduced by the plaintiff in support of its contention in that case [207]*207that this “machine” does something automatically. As a matter of fact, the above statements are incorrect as is made clear by the testimony and exhibits in the present case. The glass is not “automatically sucked into the neck moulds.” It is the action of the operator that performs this function. The bottle is not “automatically” blown by the machine. The operator manually controls the compressed air and permits it to enter the mold and blow the bottle. If left to the machine, no air would enter and no bottle would be blown. Furthermore, the finished bottle is not “ automatically” ejected from the mold. The mold is opened by the operator and the bottle removed by hand. As was indicated in the dissenting opinion in our appellate court, the original record in the Griffon case, supra, did not give a description of the “machine” involved in that case (the same machine as is involved in the present case). That defect has now been corrected and the present record gives such a complete picture of this apparatus that it makes clear what in the first record may have been in doubt, that the “machine” is not automatic in any single operation. We give the following additional definitions from Corpus Juris Secundum, Yol. 7, page 1296:
Automatic. Self-acting; or the elimination of human agency or volition, which results in the saving of labor and increases certainty and uniformity of operation; having an inherent power of action or motion, self-acting or self-regulating, not voluntary, not depending on the will, mechanical. [Italics supplied.]
and at page 1297:
Automatically. Acting without the continued application of human agency or volition, that is, as opposed to acting rationally or volitionally; of its own accord; self-acting. [Italics supplied.]
In American Roll Gold Leaf Co. et al. v. W. H. Goe Mfg. Co. et al., 212 Fed. 720, 724 (C. C. A. 1st 1914), the court said:
The word “automatically,” as applied to mechanism, is in common use and is unambiguous. It means “self-acting,” and it implies a certain cycle of movements which the machine itself makes without outside control. This cycle may be simple or complex. In the development of many machines there can be traced a constantly increasing extent of automatism; by which is meant that many steps or processes, which formerly had to be started, stopped, or controlled by the operative, are now started, stopped, or controlled by the machine itself. [Italics supplied.]
Nothing about the “machine” before us responds to the foregoing description. Not one operation was “self-acting” or made by the machine itself “without outside control.”
In a recent decision in the Supreme Court (Schwegmann Brothers et al. v. Calvert Distillers Corp., 341 U. S. 384, Douglas, J., May 21, 1951), Mr. Justice Jackson, in a concurring opinion, said in connection with the weight to be given to legislative history:
Resort to legislative history is only justified where the face of the Act is inescapably ambiguous, and then I think we should not go beyond Committee [208]*208reports, which presumably are well considered and carefully prepared. I cannot deny that I have sometimes offended against that rule. But to select casual statements from floor debates, not always distinguished for candor or accuracy, as a basis for making up our minds what law Congress intended to enact is to substitute ourselves for the Congress in one of its important functions. The Rules of the House and Senate, with the sanction of the Constitution, require three readings of an Act in each House before final enactment. That is intended, I take it, to make sure that each House knows what it is passing and passes what it wants, and that what is enacted was formally reduced to writing. It is the business of Congress to sum up its own debates in its legislation. Moreover, it is only the words of the bill that have presidential approval, where that approval is given. It is not to be supposed that, in signing a bill, the President endorses the whole Congressional Record. For us to undertake to reconstruct an enactment from legislative history is merely to involve the Court in political controversies which are quite proper in the enactment of a bill but should have no place in its interpretation.
By and large, I think our function was well stated by Mr. Justice Holmes: “We do not inquire what the legislature meant; we ask only what the statute means.” Holmes, Collected Legal Papers, 207. See also Soon Hing v. Crowley, 113 U. S. 703, 710-711.
If any ambiguity is present in the wording of the present statute, it could only attach to the question of what the Congress meant by automatic machine. To go beyond that is to create ambiguity. There is nothing in the hearings to which our attention has been directed that indicates what Congress meant by automatic machine save only that testimony hereinbefore set forth indicating that to at least one witness, and presumably accepted by all, a bottle-maldng machine was one where the work was “handled entirely by the machine * * * and not touched by human hands until it is ready to be sorted and packed.”
It would appear from the record that bottles, similar to those before us, are produced by various manufacturing methods. One method is by automatic machine, where once the molten glass is introduced, the rest of the manufacturing process is carried out by the machine. Such a machine was the subject of our decision in the Riedel case, supra. There, the machine was automatic but the raw material was fed by hand. In the present case, a graphic description of the automatic machine, with automatic feed, is shown in the motion picture (defendant’s exhibit 14). A second method is one where the glass blower, with lung power, blows the bottle in a mold. That method produces what is known as hand-blown bottles. A third manufacturing process differs from both those above described. Here, a 3-man team works together, using molds and compressed air, and manually controls each and every operation. This is the method used to produce the bottles in the present case. How many, if any, other and different processes there may be for producing bottles, we do not know. It is, however, clearly and unequivoeably stated by [209]*209the Congress that only the automatically produced bottles are subject to the lower rate of duty. We understand the provision to mean what it says, that bottles made by any method or process other than automatically by machine are subject to the higher rate. The bottles now before us were not made by an automatic process (automatic machine) but were manufactured by another method (otherwise produced) and are therefore dutiable as assessed.
We therefore hold the bottles here in question properly dutiable, under paragraph 218 (e), Tariff Act of 1930, at the rate of 75 per centum ad valorem, of at 37 % per centum ad valorem under the same paragraph, as amended by the trade agreement with Czechoslovakia, T. D. 49458, as assessed, as bottles produced “otherwise” than by automatic machine.
The protests are overruled and the classification of the collector is sustained.
Judgment will issue accordingly.