Peer review is an essential and integral aspect of Geochimica et Cosmochimica Acta. The fundamental role of the reviewer is to advise the Associate Editor and the Executive Editor on the virtues, or lack thereof, of a manuscript submitted for publication. Typically it is the reviewer who will have the most direct and expert knowledge of the field addressed by the manuscript, so that the reviewer's advice is critical, not only in evaluating whether the manuscript should be accepted for publication but also in helping to make the manuscript as useful as possible to the journal's readership.

This page provides the journal's information for and instructions to reviewers. It is divided into a number of sections, each accessible by its own link. The sections and their links are gathered into two groups, one describing the mechanics of performing a review, and one offering guidelines for manuscript evaluation. (The two group headings are also accessible by the links above.)

In general, a given manuscript will be assigned to an Associate Editor (AE), who in most cases will have the principal responsibility of identifying prospective reviewers on the basis of the authors' suggestions and his/her own knowledge of the relevant discipline.

You, the prospective reviewer, will first be asked to review the manuscript, either directly by the AE or by the editorial office. Most commonly the request will be made by e-mail, but other forms of communication may also be used, as circumstances warrant. Before you are asked, the manuscript will be posted on the editorial office's website (see Manuscript Access below), where it is immediately accessible for your inspection.

It should not take you long to decide whether or not you can do the review, so we ask for a prompt response to the request, the same day or the next business day if possible. If you need more time to reach a decision, please acknowledge the request promptly (so that we know you are not off in the field for a month, for example), and let us know approximately when we can expect to hear from you. If you agree to do the review, please plan on completing it within four weeks (see Timelines below).

In most cases the request will come by e-mail, which should also be the easiest way to respond. If it is more convenient, however, you may also respond by phone or fax.

If you agree to do the review, when you tell us that you will do the review will you please also tell us whether you will access the manuscript on the website (see below) or whether you want hardcopy in the mail. If you don't say anything either way, we will have to send you another message asking that question explicitly.

If you cannot do the review yourself, but can suggest someone else who is qualified, we will welcome your suggestion, particularly of someone, perhaps working in your research group, who is a young postdoc or senior graduate student, who might not yet be well known to editorial staff. But please just make the suggestion to the editorial office or the AE, and do not actually transfer the request to your suggested prospective reviewer (by giving him/her the access codes and/or a copy of the manuscript file).

All manuscripts will be posted on this (the editorial office's) website. To access a given manuscript, click the Manuscript Access button on the navigation bar at the top of most pages. On the Manuscript Access page there will be a textbox into which you should enter the five-character manuscript number "Wnnnn" for the manuscript you want to access (where "nnnn" is the appropriate four-digit number). You will then be presented with a dialog box in which you are asked for a username and password. Both of these will have been supplied with the review request. Normally the username will be "r-Wnnnn", where "nnnn" is again the four-digit manuscript number, and the password will be a six-character string unique to reviewers of that manuscript.

Please note that the recognition software for manuscript access is case sensitive, so you have to enter the information exactly as illustrated above or supplied with the review request.

After correct access code entry you will be presented with another page specific to that manuscript and containing links to the files associated with it. There will always be a file for the whole manuscript, double-spaced. In some cases there will also be a single-spaced version. There will also be a file for title page and abstract only. You may access any or all of these files. If you have a very slow connection, you might want to access just the shorter title-and-abstract file if you can decide whether to do the review on that basis, and, if you agree to the review, access the full manuscript later.

Sometimes there will be still other files on the manuscript page. Possible other files are otherwise-inaccessible papers/documents cited in the manuscript you are reviewing (e.g. see Self-Referencing below), and/or (if you are being asked to do a re-review) one or more generations of revised manuscript.

Each item in the list of files is a link. When you click on it your browser will either attempt to open it or download it to your computer, depending on how the browser is set (often, a right-click will offer you the choice). Either way, you can examine the manuscript directly on your computer or print hardcopy if you wish.

At the bottom of the manuscript page, below all the file links, you will see another link, labeled Reviews and Calendar. This link will lead to another group of files associated with this manuscript, but intended for access by authors only, not reviewers. If you click this link you will again get a dialog box asking for username and password, but your username and password will not work, since you don't have the username/password given to the authors.

Manuscript files are in PDF format. To open them you will need the Adobe Acrobat Reader, version 4.0 or higher (earlier versions sometimes work, but sometimes don't). If you don't already have it, you can download the current version of the Reader for free, from the Adobe website (www.adobe.com).

If you lack internet access, just let us know and we will fax or e-mail you the title page and abstract to help you decide whether you can do the review.

If you agree to do the review, we hope that you will access the manuscript electronically, i.e. by downloading from our website and, if you like, printing a hardcopy yourself. If you prefer, however, on request we will send you a hardcopy in the mail. Either way, please let us know if you will access it electronically or you want hardcopy when you tell us that you agree to do the review.

Reviewers are reminded that manuscripts submitted to a scientific journal are confidential documents until they are published. The manuscript you are reviewing or are asked to review should be discussed only with the AE or with editorial office staff. Your access to a manuscript under review is privileged; please do not disclose its content (or even title or authors) to any other individuals except with specific approval of the AE or the editorial office.

Of all the factors that contribute to the length of the review and evaluation process, the greatest is usually the time a manuscript spends on a reviewer's desk, or in a reviewer's briefcase, waiting for him/her to get around to doing the review. We ask for your cooperation and rely on your good will in not prolonging this interval to no good end. The Golden Rule applies: It is not unlikely the authors of the paper that you are presently reviewing will someday be reviewing one of your own papers; review as you would like to be reviewed.

The general guideline is that you should complete your review as promptly as is reasonable. More specifically, we ask that you complete your review within four weeks of the time you agree to do it. If specific circumstances warrant, this may be extended to six weeks, but this should be uncommon and should be discussed beforehand with the AE or with editorial office staff. Please be realistic: If you anticipate that you cannot follow these timelines because of field work, teaching schedule, proposal deadlines, etc., don't agree to do the review in the first place. If you promise to do a review and then fail to deliver, the likely result is that your review will be abandoned and the evaluation process will have to proceed without benefit of your review. This will serve no useful purpose and will instead serve only to annoy all parties concerned.

If you review becomes overdue, you will get a reminder message from us. Please do not take offense at this. Sometimes reviewers really do forget, and the message serves an important literal reminder function. Sometimes the message serves as a useful goad in cases in which the reviewer had not forgotten but was only procrastinating. The reminder messages also serve the critical function of establishing a real deadline. We understand that sometimes a reviewer simply cannot carry out a promised review because of unexpected professional responsibility, health problems, family matters, etc. If this happens to you, please tell us, and we'll move on. If we don't hear from you, however, we will send you a message that says we will wait (only) one more week for your review, and if we don't hear from you in that week we really will abandon your review and move on without you.

Please submit your review on the editorial office website. Click the Manuscript and Review Submission button on the navigation bar, then, after you read the instructions on the resultant page, click the Review Submission link. You will be presented with a dialog box which requests username and password. Use the same username and password that you used for access to the manuscript on the wesite (and remember about case sensitivity).

After entry of the correct username and password you will be presented with a "review form" submission page, which will include textboxes and scrollboxes. Section 1 on this page will display information which identifies the manuscript for which you are about to submit a review (which is implicit in your username/password). In the first two textboxes (Section 2) please enter your name and e-mail address. Then (in Section 3) there are four scrollboxes in each of which you are asked to select one of several options; this is not the most important part of your review, but it provides a useful summary.

Also in Section 3 there is a textbox, under Reviewer Comments, into which you are invited to enter whatever you have to say about this manuscript (but see below). This is the most important part of the review. There is no checklist of points to cover; say whatever you think is important to say. You may type text directly into the textbox, or paste from another program.

At the end of Section 3, below the Reviewer Comments textbox, there are four small textboxes in which you can designate files to be uploaded from your computer to our server (it is usually much easier to use the Browse button associated with each box than to enter the exact file specification directly). Uploaded files (text, figures, tables, whatever) will be part of your review. The file upload facility works very much like the manuscript submission upload facility, and it is also subject to similar restrictions on file type and length and on filename extension. Allowable file types (indicated in filename extension) are listed near the boxes.

After you have entered text, selected scrollbox options, and perhaps specified files to upload, click the Proceed to Proof button at the bottom of the page. This will take you to a proofing page, from which you can either submit the review (the Submit Review button at the bottom) or return to the review form page (the Edit button) to make changes.

If you click the Edit button, and go back to the review form page, the information you entered in the name, e-mail address and reviewer comments textboxes, and your scrollbox selections, will all be preserved. The upload-file specifications will be lost, however, and will have to be re-entered. You should move from one page to another only by using the navigation buttons provided. If you use the Back, Forward or Reload/Refresh buttons on your browser you may lose all the information you have entered. (You may also lose all information entered if the process takes too long, more than a few hours.)

When you click the Submit Review button the text your have entered, your scrollbox selections, and any optional files that you designated, are uploaded to our server and constitute your review. You will be presented with a page that notifies you that "Your review submission is complete" and gives you a "tracking number" for reference. (If you do not see this message, and get a tracking number, you have not submitted a review yet.) This page also provides a link back to the homepage.

Although you can enter or paste whatever text you like, of whatever length, into the Reviewer Comments textbox, we suggest that the main body of your narrative comments and recommendation is best incorporated into a separate file and uploaded rather than entered into the textbox. This approach will not only allow you to keep a copy for your own records, and mitigate any lost work in case you should lose your internet connection, etc., It will also give your better control over the format in which your comments are seen and it will allow you to use document-preparation features (variable fonts, bold and italicized text, sub/superscripts, special symbols, expanded alphabets, etc.) that are not available in a textbox. Such an uploaded file is commonly a word-processor (Word or WordPerfect) file but may also be a .txt, .rtf or .pdf file.

Although our preference for review submission is website upload, we will be pleased to accept your review by other vehicles that suit your convenience. one possibility is e-mail, with main narrative comments either in the body of the e-mail or in attached files. You may also submit your review by fax or in hardcopy (courier or snailmail). (See homepage for address and fax number.)

If your preference is hardcopy, you can print your comments on plain paper or on your own letterhead. You can also use our hardcopy review form. A paper copy of the review form will be included with the manuscript if you asked for hardcopy, and/or you can download from this website a pdf file from which you can print the review form or editable word processor review-form files which you can use as templates (see links on the Information Page).

For your information, whatever material you submit as part of your review, in whatever format, will be assembled into a pdf file and posted on our website, generally on the same day it is submitted or received. The posted review file will be immediately accessible to the AE and it will be made accessible to the authors at the time of the AE Report. Also, however we receive your review, we will send you an e-mail acknowledgment message to let you know we got it.

Reviewers sometimes want to write, by hand, directly on the (hardcopy) manuscript, either to make minor comments pertinent to very specific parts of the text (correction of typos or spelling errors, recommendation to substitute one word or phrase for another, and so on), or to flag locations discussed in the narrative review. If you want to do this, that's fine; send the annotated hardcopy manuscript to the editorial office and we will scan it and incorporate it into your posted review so that the AE and, in due course, the authors, will be able to see it. When the AE report is transmitted to the authors we will also mail them your annotated hardcopy. If you want to make annotations on a hardcopy manuscript, please use a pen (pencil does not scan well) and write boldly. Please don't annotate in red ink; red scans OK in color scan mode but not black-and-white scan mode, so red annotations require much larger file sizes.

As an alternative, you can electronically annotate the manuscript as the pdf file you download from our website. Adding annotations to a pdf file requires the full Adobe Acrobat program, not just the Reader, but if you have the program and can make annotations this way, they can be quite effective, and we welcome manuscripts annotated electronically. This approach has the advantage that authors can read typescript rather than having to interpret handwriting, and also that the annotated manuscript is an electronic file which can be transmitted instantaneously rather than requiring snailmail.

The default situation is that your review is anonymous, i.e. your identity is known to the editorial office and to the AE, but not to the authors. In communicating with the authors we will not mention your name to the authors unless you have have already done so yourself. If you wish to "sign" your review and let the authors know who you are, write your name somewhere in your main narrative (the Reviewer Comments textbox on the website, or in an uploaded file, or in a hardcopy review.

If you send in your review in hardcopy, and you use your own letterhead, it will be inferred that you have signed your review (we will scan the hardcopy, so the authors will see your letterhead); use plain paper if you don't want to "sign" your review. If you fax your review to the editorial office, we will blacken out the line with the originating fax number (before scanning it) so that the authors cannot identify you in this way.

Many of the files (such as word-processor files) that you create on your own computer will include your own name (or your log-in username) as "author" of the file, and this information can be revealed by examining file "properties". Your name will remain associated with the file properties when you send it to us, whether by website upload or e-mail attachment. In general, you can suppress inclusion of this information in the program that creates the file, but this is not necessary, even when you intend to remain an anonymous reviewer. When we make a pdf file from your file, your name information will be replaced by that of office staff. If you send us a pdf file, before posting we will print it to the Adobe Acrobat distiller and make another pdf file, again overwriting author information.

Feel free to communicate informally directly with the AE, either at his/her instigation or yours. If the communication is in some black-and-white form (e-mail, fax, post), please copy it to the editorial office. If by phone or in person, there is no need for a transcript, but it would be useful if there were some black- and-white allusion to the conversations. Your final, formal review, however, should still be submitted to the editorial office, not directly to the AE.

Peer review of a scientific manuscript is based on judgement, and application of evaluation criteria is typically quite subjective. Most experienced investigators usually have definite ideas about how to do a review, in general as well as for Geochimica in particular, and don't need to be told how to do it. Still, it wouldn't hurt to have a look at the following guidelines.

Despite the authors' best efforts, manuscripts are often not free of problems such as typographical errors, misspellings, minor grammatical errors, incorrect figure or table citations, missing or superfluous references, etc. Reviewers are wont to point these out, and doing so is a service to authors, editors and readers.

Rarely, but not never, the frequency of such mechanical errors rises to a level at which it is both an annoyance to the reviewer and a distraction from the scientific message. If you consider that this is the case, or more generally that the authors have been careless or negligent in preparing the manuscript, don't do a scientific review, just state your opinion that the manuscript is "unfit for review". This should be automatic if the manuscript has a missing figure or table.

Irrespective of intrinsic scientific merit, a manuscript is not suitable for publication in GCA if its subject lies outside the journal's scope. If it is your judgement that the manuscript in question is inappropriate for GCA because it's outside that scope, then you can just say that without reviewing the science.

You can look up GCA's statement of scope if you like, but it's not a complicated statement. The scope of Geochimica et Cosmochimica Acta is geochemistry and cosmochemistry. In most cases there will be no ambiguity about whether a given manuscript is within the journal's scope. In marginal cases about which you may be unsure, say that it's marginal and/or that you're not sure about it. You should also feel free to discuss appropriateness with the AE or with the Editor, by e-mail, by phone or in person. In a very real sense, the definition of just what constitutes geochemistry and/or cosmochemistry is in the perception of the scientific community that practices these arts and answers to those names, so your opinion on this issue is important.

A manuscript might depart from GCA's scope in many different directions, but in practice it seems that the most frequent departure is in the direction of a manuscript which is about chemistry but not necessarily about GEOchemistry. Remember that it is the authors who have the burden of proof to demonstrate that their manuscript lies within the journal's scope. If you consider that the authors have not successfully carried this burden of proof, but maybe they could, you might want to recommend that they be asked to rewrite, adding a discussion of appropriateness within GCA's scope, for ab initio review.

But note, however, that GCA's scope is not restricted to natural geochemistry. We have ample precedent for publishing papers whose focus is pollution, mine wastes, radioactive waste containment, etc., areas in which human activity is involved.

In evaluating a manuscript you should also consider its significance and its relevance to GCA's readership and to the geochemistry/cosmochemistry scientific community at large. Perhaps someone has analyzed the abundance of thirty elements, and the isotopic composition of half of them, in a MORB or in a randomly selected roadside pebble, and done exquisite analytical work, and provided a thorough description of experimental procedures and an honest discussion of results. That's geochemistry, but unless there's a strong tie-in to MORBs (or pebbles) in general, or how they're formed, or the evolution of the mantle, or something of broader relevance, that's neither very significant nor very interesting to readers, and it should not be published in Geochimica. Similarly, manuscripts dealing with topics of only local interest, without broader significance clearly spelled out, should not be published in Geochimica. If there is any question about the broader relevance and significance of a very narrow or local study, it it the authors' responsibility to make the connection, not the reviewer's.

Besides the actual technical issues discussed in a scientific manuscript, the expository prose style in which a manuscript is written is also important in getting its message across. Criticizing expository style - as well as grammar, punctuation and spelling - is certainly fair game in the review process.

Geochimica manuscripts are required to be written in English, but it is an international journal whose authors represent many different countries, in most of which English is not the native language. Understandably, some of these authors find writing in English to be challenging. We ask for your tolerance and understanding in such instances. Perhaps the prose is less than elegant, and perhaps frequent grammatical mistakes are distracting, but please try to concentrate on the science as long as the authors' meaning is clear. If the meaning is really not clear then there is little choice but to reject the manuscript and perhaps ask for a new one, written more clearly (e.g. by enlisting the help of an English-fluent colleague), to be reviewed again. If the meaning is clear, even if awkwardly expressed, and the science is good, so that ultimate publication is to be expected, you can help by making corrections and suggesting sytlistic changes (a common use of manuscript annotations). Fixing the English is not really a reviewer's responsibility, but authors are frequently very appreciative of any linguistic help and it can significantly improve the scientific utlility of the published manuscript.

It is natural enough that authors commonly reference their own prior work, or that of their research groups, rather extensively. Sometimes this leads to a problem in reviewing a manuscript, however, if authors reference their own work for something important in understanding or believing what they say in the manuscript you are reviewing (data, conclusions, discussion, description of procedures, sample or field descriptions, etc.) and the reference cited is not accessible to you, perhaps because it is unpublished or published only in gray literature (e.g. a student thesis or conference abstract) or perhaps because it is not yet published (in press or submitted). If that's the case, and you want to see the referenced work in order to evaluate the manuscript in hand, tell the editorial office about it and we will request that material from the authors. When we get it we will typically post it on the website, alongside the manuscript under review, so that other reviewers as well as you can access it. If the authors are unable or unwilling to provide such material, we will either reject the manuscript in hand or suspend evaluation until such time as the paper(s) in press or submitted are published and accessible.

If the authors cite their own work that is "published" only in a way that you think will be hard for most readers to find or access (e.g. a conference abstract or thesis that is not publicly accessible), and you think it's an important (for understanding or belief) reference for the manuscript you're reviewing, please call attention to this circumstance in your review. If the manuscript is otherwise acceptable for publication in GCA we may ask or require the authors to provide the obscure work as an electronic supplement for the internet version of GCA.

We have no qualms about authors referencing their own papers in press, as long as reviewers and editors can get access to them (see above). The same is true, during the initial evaluation process, for cited papers which are listed as submitted. Our policy, however, is not to commit to print a manuscript which references another manuscript as submitted. Instead, we will hold up publication of a manuscript until the other manuscript changes from submitted to in press, i.e. until it is accepted for publication. If the cited manuscript is not accepted, our manuscript, the one which cites it, will have to be rewritten so that it is no longer necessary to cite the manuscript which was rejected. If the manuscript you are reviewing contains a substantive reference to another manuscript which is submitted, it would be useful if you called attention to this fact in your review, to make sure that a submitted reference does not slip by.

A manuscript which is referenced as in preparation is something else again. This is a manuscript which they authors are only planning to write, or perhaps just thinking about. Often, a manuscript listed as in prep in short order becomes submitted and then in press, but all too often it is never completed at all. Even if the in prep manuscript is ultimately written and published, an in prep reference in the journal does the reader no good. Accordingly, in prep references are not permitted in GCA. If you encounter an in prep reference in the manuscript you are reviewing, please call explicit attention to it in your review, and offer advice on how important it is and what should be done about it. If the in prep reference is of only peripheral interest or marginal utility (and the paper is otherwise satisfactory), we will require the authors to eliminate the reference in revision. If you think that the reference is important to the manuscript you are reviewing, e.g. it is supposed to contain pertinent data or describe analytical procedures, such that you might need to see it (as above) if it existed, we can reject the manuscript you are reviewing and decline to consider a resubmission until the in prep paper becomes a reality.

There are several more-or-less objective criteria involved in manuscript evaluation, as described below. In your review it is not necessary to check off each one; typically reviewers will mention these various aspects only when something is wrong with them. This is not inappropriate; sometimes an identified flaw will prove fatal, but more often remediation of a flaw helps improve the paper.

Foremost is the issue of whether the authors have made some substantive scientific mistake (in the reviewer's opinion, anyway). Perhaps the data, or their error bars, are suspect. Perhaps the authors have made some mathematical mistake in data reduction or in derivation. Perhaps figures are misplotted. Often enough, if more subjectively, you may consider that there is something wrong with how the authors proceed from premises and observations to conclusions and generalizations. In most cases it will have little effect to ask that the authors change their conclusions to alternative interpretations more favored by the reviewer, and rightly so, since it is the authors' names on the paper, not the reviewer's. But it is certainly appropriate to ask that the authors fill in intermediate steps in their reasoning and/or that they at least mention alternative interpretations.

Reviewers should also consider whether the work is original, and/or whether the authors may have reported substantially the same thing elsewhere (exclusive of abstracts). A related question is whether the authors have cited and given credit to relevant antecedent work by others. Priority disputes are usually ugly and distasteful, and it is better to avoid them at the outset by calling attention to any perceived deficiency in giving credit where credit is due.

There is no arbitrary preset limit on the length of a GCA paper, but that doesn't mean you should ignore length. If you think a manuscript is too long, in terms of tables and/or figures as well as text, and can be shortened without serious damage to the message, please say so. The issue is not just saving trees. Everybody's busy and, other things being equal, more people will read a short paper than a long one, and you will serve readers well if you help minimize dilution of the science by unnecessary figures, tables and/or verbiage. But don't just object to length as a matter of principle: Say which figures or tables can be dispensed with, or where text should be trimmed or eliminated.

The converse also applies. If you consider that science will be served by adding tables or figures, or adding prose to describe methods, reasoning, implications, etc., say that as well.

Reviewers sometimes seem to think that their job is to identify what's wrong with a manuscript and tell the authors to fix it. That's true, but it's only part of the story, Even if there's nothing wrong with a manuscript, or if whatever's wrong can be fixed, that doesn't mean it should be published in Geochimica. To warrant publication, there has to be something right with it too. The work should be of high scientific quality, it should be original, it should constitute a significant contribution to scientific understanding, and it should be something that the readership will be (or should be) interested in (see Significance and Relevance). Just how high the bar should be is a matter of judgement and impossible to define in general terms; this is another respect in which reviewers help determine the nature of the journal and the state of the art.

So please don't confine your review to identifying what's wrong with the manuscript. If your opinion is that the manuscript warrants publication, say why - say what's right with it.

Sometimes a review discusses many aspects of a manuscript without providing a clear bottom line. Make sure the Associate Editor and the Editor know whether you think the manuscript should be accepted or rejected or re-evaluated after certain changes are made. Even when you think the intrinsic scientific merit of the manuscript warrants publication, you may make several recommendations for changes; if so, make it clear which are just constructive suggestions that the authors may or may not adopt, at their discretion, and which you think are necessary before the manuscript is suitable for publication. In general, the reviewer's principal function is to provide expert advice, so you should make it explicitly clear just what your advice is.