to affect the voluntariness of the confession. In the questioning of a witness or suspect an unobtrusive microphone will pick up all that a stenographer will, but when the witness is talking freely he will not be brought to a halt and lose his trend of thought by someone saying "You'll have to go slower, I can't get what you're saying"!

I have witnessed captain's mast at a busy receiving station, and I was convinced that a mechanical record of what transpired would be very useful for the personnel yeoman and the discipline officer. Although all records of the "desk" in each shore patrol headquarters are now kept either in longhand or by filling out forms, a mechanical record. of examinations would protect SP personnel against charges of maltreatment. Further, they would be evidence of what the offender and the witnesses said at the time when the details of the offense were fresh in the minds of all.

It is a noteworthy point that in many cases these records could be stored for a period before being cleaned for re-use. There is no need to transcribe records to typewritten form when they will, for most purposes have the same practical value if listened to. Typewritten transcripts are necessary only as to that part of the material which must be sent to someone else.

As a result of the above, and to examine further into the subject, I gathered together some information on the various kinds of recorders, and made some experiments with a few of them to see how they worked out in actual practice. But first, for the information of those not familiar with general court martial procedure, let me explain the usual system of reporting these courts as it is now practiced. The reporter sits in the courtroom and makes notes in shorthand of all the evidence and procedural steps, but what he does with them after the session is not so well known. The reporter reads his notes, and from them dictates all that has happened in court. This dictation is done into a cylindrical recorder. The records of this machine are then transcribed by clerks, sometimes to a rough, but usually to the smooth, finished copy. The reporter then compares this copy with his notes and, if satisfactory, he turns it over to the judge advocate as the record of proceedings. Under this system, a court reporter can use his highest skill, i. e., speed in shorthand, only about 15 hours in a 40-hour week. The rest of his time must be spent in dictation and editing.

The first mechanical system considered, which

we will call system "A," eliminates two steps in the customary procedure described above. The reporter takes no shorthand notes in a book, but instead he repeats or dictates into a speaking mask covering the lower part of his face whatever is said by the parties to the trial. The mask makes his voice inaudible in the courtroom, but he tells his machine the words spoken by the court, the witnesses and the judge advocate, adding identifying matter, descriptions of exhibits and gestures if used by witnesses as part of their testimony. The reporter's voice is recorded on a disk mounted on a turntable which will record sound for about 30 minutes on each side.

I did not experiment with this system so I cannot say whether there are technical faults in the equipment or machinery used. From a clerical viewpoint, the reporter is dictating directly from the scene into the record so that he eliminates his shorthand notes with all their imperfections and limitations. When the records are transcribed by the typists, the records are the actual impression of what the reporter heard and saw in the courtroom, not the record of what he read from his notes a few hours or days later. The system has the advantage of saving the reporter for work in the courtroom by not using his time in reading over his notes and dictating them into cylinders. Then too, as each hour's work is done, the disk record can be removed from the turntable and can be sent to the transcribing room, so that the record is made up as the trial continues.

In considering this system, I felt that it would be difficult to train a person to sit in court, to listen to a conversation and instantly to repeat it into a machine. The sense of hearing requires time to hear the speaker and understand what he is saying. After that it is necessary to impel a motor reaction so that one's vocal mechanism will repeat what is being heard. I had one man read aloud from a book while I tried to repeat what I heard into my cupped hands. The result was confusing, as I found I had to stop talking to listen, and to close my ears while I talked. However, as the reporter pointed out, it should make little difference to a trained man what his motor reaction is. If he is trained to hear speech and to have conditioned reflex actions in the form of handwritten stenographic forms, his reflexes could be just as easily conditioned to react to conversation by repeating

it.

This "A" system has been tried in a busy naval

district, with complete success. It took only 2 weeks of training before the reporter felt confident and competent to use it, and the results were satisfactory both from the viewpoint of the judge advocate and from the aspect of costs. It resulted in a marked reduction in the time lag between the trial and the submission of the completed record of proceedings. This system is also used with success by many contract reporters for general speed reporting of hearings and Government agency matters.

I tried extensive experiments with a wire recorder but they ended in failure. A recorder was used that cost the Government over $600. This device used a microphone that picked up the sound and carried it to an audio amplifier, which in turn carried it to a magnetic recording head. Through this head a thin, fine wire passed from one spool to another and, as it passed through, the sound was recorded on the wire. I gave this machine a thorough work-out over a period of 2 months and then returned it.

This machine was installed in the courtroom to take the record concurrently with the reporter. As the machine was portable, though very heavy, the temporary installation was inexpensive, and it had numerous other assets. When the wire record was rewound and played back, all sound in the room was recorded and voices were faithfully reproduced and easily identifiable. If a person could be understood in ordinary conversation, he could be understood on a play-back of the wire. By putting the microphone unobtrusively in position and by not referring to it, and by placing the recorder and its operator behind the witness and out of his sight, there were no cases of "mike-fright." The courtroom was airconditioned and had sound-absorbent covering on the ceiling, so that there was practically no interference with the recording of voices. In other rooms where there was no sound-damper, some foreign noises crept into the record, but not enough to cloud the voices.

However, so many faults developed in wirerecording that it is clear that considerable improvements must be made in the machines before we can rely on them for a sure record of proceedings. The machine itself, although expensive, was mechanically very faulty. The motor had a dry clutch which could be switched either to pick up the sound or to rewind the wire. The motor required oil, and a drop of oil got on the clutch

and destroyed its friction, causing the spools to wind at an irregular speed. The machine ha several wheels revolving on shafts, the bushings of which required oil. When so oiled, the revolving motion threw off the spare oil so that the belts operating the wheels became slippery and funetioned irregularly. Where the shafts had no oil. or very little, they soon became heated and bound at the bushings so that they had to be dismantles and polished before they fitted again. Then too, if the electric current were irregular or low in voltage, the speed of the machine dragged. All these faults garbled the record beyond recognition.. Frequently the record was garbled before the operator knew it, so that much of the testimony was improperly recorded or not recorded at all and, had not a regular reporter been taking the proceedings, the court would have been without a true record-a position hard to explain.

The above difficulties were encountered in experiments with only one machine of its type. The design of any machine must be such that a positive, geared drive is provided, to revolve the spools at constant speed. Any difficulties of lubrication, binding shafts, slipping clutches and belts should be engineered out of the machine before it is built. if it is to be reliable and valuable in Navy legal proceedings. The machine should provide a more rapid rewind and play-back of the record. At present, with the recorder tested, if a proceeding has been under way for 35 minutes and you want to repeat something that happened at the fifteenth minute, you must reverse the machine and rewind for 20 minutes. Then you can hear what was said. but you must again advance the wire for 20 minutes, until you come once more to the thirty-fifth minute and take up the trial where you left off. Our play-back at this point has cost us 40 minutes. This undesirable feature does not exist in disk recording where all that need be done is to move the arm back across the disk to the point under inquiry and turn the switch to play-back. The highly desirable feature of rapid play-back should be insisted upon in any machine adopted for general use.

Other wire recorders were examined. The less expensive and less elaborate they were, the more sturdy and trouble-free they seemed to be. All, however, appeared to have the same inherent defect of snarling the spools of wire, and I saw none which had any device to insure against snarling and breaking. The wire is so thin and fine that

when it breaks jeweler's tweezers must be used to hold the ends while joining them. The wire is wound on the spools very carefully, to avoid tangling, but, when tension is lost on the wire, the wire loosens. Then all the various coils intertwine and mix up in different planes, so that all the top coils of the spool become a snarled mass of knotted and tangled steel wire. Because untangling this is impossible, it must be cut out and the remainder spliced. Should a snarl or kink remain undiscovered in the lower coils on the spool, it will cause trouble, perhaps even breaking again and causing a new snarled mass to form. Such a spool is soon not usable. If you have a recording on it, that part which is cut out is lost to you-a loss which could be very serious if it were the only record of an hour's testimony before a general court martial.

I have heard good reports about tape recorders, but was unable to get one for experimental purposes. These are similar to wire recorders, except that they use a paper tape instead of a wire. In the tape is imbedded iron or steel dust which has the same recording capacity as the wire and about the same, or higher, fidelity. Paper tape has had wide commercial use over a long period, as in stock tickers and telegraph instruments. Its use as a voice recorder is new but it offers good possibilities. It is cheap, and if it breaks it is easily mended with cellulose tape, and because it will not snarl there is no loss of the record. There are quantities of literature on the advantages of tape recorders, which make great claims for them. They are instruments with great potentialities and merit considerable further study.

Others in the Navy have conducted similar experiments during the past several years. Inquiry brought to light the successful use of devices that record sound on film which, as we all know, gives high fidelity sound reproduction. Also in use, successfully, are disk recorders of the type used in broadcasting studios. In general these two types of recorder seem to be too expensive, too technical, and too complex for operation by personnel who have limited training.

I know that we tend to resist mechanization. We want to adhere to the familiar, the old way of doing things. The cowboy has not given up his horse, the yachtsman allows no substitute for the thrill of sailing and antique shops do a thriving business. But we have added central heating to our colonial homes, the oil lamp is out and the automobile is here to stay. The illustrations of

improvements are endless, as more economical ways of doing things appeal to our good business sense. Slowly, even the most reactionary ceases his resistance and permits the change.

The Navy needs good mechanical recorders. A workable instrument should be developed for use in the general service. To operate these, personnel should be trained, not only in the mechanical features, but also in the forms and procedures of our courts and boards. When this is done we will have better and less costly recording.

From my observations and inquiries I came to several conclusions. It takes less time to train men to operate mechanical recorders than to learn. stenography, so that we save from that aspect. Mechanical recording is faster than stenography, and as there are fewer steps and no dependence on memory, there is less margin for error. It is more economical than stenography and should produce highly efficient results with less cost for clerical salaries or other overhead. It has great mobility and can be used afloat or ashore. It need not be fed or berthed, will never go over the hill, and is guaranteed not to ask for a chit to the chaplain.

ADMIRALTY

THE HE COST of the operation of naval vessels lies not alone in fuel oil, supplies, wages, and subsistence of naval personnel, etc., but also in the inevitable aftermath of the claims for damage which grow out of their operation.

Prior to 1944 the Navy Department could effect settlement of vessel damage claims only through recommendations to the Congress where the amount involved was not in excess of $3,000. After the Secretary's aproval of this limited settlement authorization and Congressional appropriation, final payment was made through the General Accounting Office. The limitations on the authorization and the time element rendered litigation inevitable, in most instances. This procedure was obviously inadequate to deal with the tremendous mass of claims arising out of the Navy's greatly expanded fleet and its wartime operations.

Accordingly, legislation was obtained which, in addition to conferring on the Navy Department security control over admiralty litigation involving naval vessels during war, afforded the Navy Department authorization to effect settlement

EGAL OFFICERS, who have to deal with set

LE

tlement of personal injury and death matters, | will be interested to know of the appearance of the NACCA Law Journal. This is the official publication of the National Association of Claimants Compensation Attorneys. Each volume contains a discussion of personal injury cases in admiralty law, workmen's compensation as well as Tort Claims Act matters. The bound journals present the matter from the viewpoint of the claimants. Three volumes have been issued. The subscription, which covers the cost of two books annually, can be entered with the NACCA Law Journal, 6 Beacon Street, Boston 8, Mass.

30 June 1946.

30 June 1947.

30 June 1948_. 30 June 1949_.

Total

1, 679, 554. 74

Most major collisions that occurred prior to July 1944 had found their way into litigation before the act was passed, and could not be dealt with under this settlement authorization. These statistics do not include settlements or payments of judgments in litigated matters since such payments are made by Congressional appropriation after the entry of final decree.

Due to the admiralty practice of mutual fault liability, the over-all statistics do not represent the total extent of damage inflicted and recoverable by opposing interests. The amount paid is reduced by the recoverable damage of the naval vessel involved in the collision.

After enactment of the admiralty vessel damage claims act, an issue arose as to the extent of legal authorization to settle affirmative claims which were the counterpart of the 1944 act, that is, claims where the Navy Department was on the collecting end. This issue was put at rest by legislation obtained in December 1945 (34 U. S. C. 600a). Since that time $702.288.38 has been collected. The experience over the several years is:

Of greater interest to naval officers is the proposed revision of the Rules of the Road which received so much publicity through the activity of the committee of which Capt. R. F. Farwell, USNR, acted as chairman. The Convention's action in this connection is Appendix 5 of the Convention, entitled "Regulations for Preventing Collisions at Sea" which appears at pages 173-187 of the State Department's publication. The regulations were not made a part of the Convention. While approved by the Convention, they will require separate action by each adhering country. The U. S. Coast Guard has under consideration the manner in which the proposed new regulations will be presented for action by the Congress.

STUD

TUDENTS of admiralty law have frequently commented on the lack of any detailed American treatise on collision liabilities. This deficiency has been remedied by the appearance in 1948 of an exhaustive 946-page treatise "The American Law of Collisions." This book was written by the late John Wheeler Griffin of Messrs. Haight. Griffin, Deming and Gardner, New York City. It represents Mr. Griffin's lifelong experience with collision litigation, based upon his research over a very extended period of time. This volume can be obtained at a cost of $20 from American Maritime Cases, Inc., Baltimore. The editing is by Arnold W. Knauth, an editor of American Maritime Cases.

16

U.S. GOVERNMENT PRINTING OFFICE: 1949