Although the large Spring Number is not with us, we are attempting once more to produce a twelve page  B.B.  This month, we take advantage of this increase in size to print a long article taking up most of the B.B.  To these of you who are not interested in cave surveying, we apologise and hope that the next time we print an equally long article, it will be on a subject you are interested in (within reasonable limits, of course!)

This article, as well as being long, is of a controversial nature, as it suggests a modification, to a well established practice.  Whenever the B.B. has printed articles which were felt to contain some controversial elements in the past, these have always performed like damp squibs, and not raised a single voice of agreement or dissent.  We hope that, in this case, we shall get some correspondence  agreeing or disagreeing with  the article, as the subject of cave surveying is one in which there may well be considerable scope for new ideas.



Whitsun weekend. There will be a club meet at Gaping Ghyll.  A coach is being arranged by Brian Prewer.  Anyone interested should get in touch as early as possible.  It was suggested that the club should visit Lancaster Pot, but this has proved impossible owing to the grouse season, approximate cost 35/-.  Date June 10th.

Some (Controversial?) Thoughts on Cave Survey Gradings.

by  Bryan Ellis

This article in no way tells you how to make a cave survey.  It deals only with one aspect of the completed survey that of applying; a grading of expected accuracy.  It is important to remember that the views expressed are solely, as far as I know, those of the author and must not be taken as representative of those of the B.E.C. committee, nor of the editor, nor of any other member.  The purpose is to express on paper some of the views of the author in the hope that they will provoke discussion.  Now that you have been warned, here goes.

Some form of survey grading is very desirable so that, by simply looking at this figure, the reader may make a reasonably reliable estimate of the accuracy to be expected. However, at the same time it is even better if accompanying each cave survey published there is an article describing the instruments used in making the survey; how the figures are calculated and the survey plotted; a list of known closure errors and so on.

Let us take four hypothetical cave surveys, those of Axbridge Hole; Burrington Cavern; Cheddar Sleeker and Draycott Swallet.  A beautifully produced survey of Axbridge Hole is published without any accompanying screed!  The survey looks very good but even the closest examination fails to reveal any sign of a grading.  It is therefore impossible to arrive at any estimate of its accuracy.  It is hardly likely that a   survey would be produced as elaborately as this for anything less than a Cave Research Group grading of 5, but one cannot be sure.  Perhaps I may know the surveyor and therefore know the instruments that he uses and can guess that the accuracy is perhaps between the grades of 5 and 6.  But this is only guesswork and if I don't know him, I would have no idea at all.

Our second hypothetical survey has been published, in a club journal; this of Burrington Cavern.  Once again, close scrutiny of the map fails to show any sign of a grading.  This time, however, the situation is much better because there is an excellent article by the surveyor accompanying the survey.  This article describes in great detail the instruments that were used, the way in which he made the survey, calculated the position of each station - and he also gives a table of known closure errors.  This article is a model of what such an article should be except that even in the text there is no indication of the grading claimed.  Possibly the idea is that, by giving the reader all the details, he should be left to form his own conclusion, but before he can do so he must read all through the text.  He can get no idea from the survey itself.  Furthermore, the surveyor himself is the best judge of the accuracy to be expected and if he gives an opinion then perhaps the reader may like to adjust it slightly after reading the text.

The third survey is of a cave known as Cheddar Slocker and, like the first, is a well produced sheet but without any accompanying description.  This time, examination of the survey shows that the surveyor has claimed a grading if 5.  Anyone looking at this map now knows that, as a minimum, a calibrated prismatic compass; metallic tape and a clinometer were used.  It is therefore safe to assume that the survey is accurate enough for most purposes.  However, as there is no accompanying article, there is nowhere to state what instruments actually were used, nor of the closure errors that were found in the closed traverses in the cave.

Finally, we can consider the survey of Draycott Swallet, our final hypothetical survey.  This survey has been published in similar form to those of Axbridge Hole and Cheddar Slocker, but this time it has been sold together with a small booklet.  On the sheet is found the grading claimed by the surveyor and therefore an estimate of accuracy can be made immediately.   In the booklet are found all the details of the survey and its making, so if anyone is interested, they can read through it and then they should have an even better idea of the accuracy of the survey.

The idea of quoting at length the details of these surveys is to give the reader an idea of why I consider both a grading and an accompanying screed to be the ideal.  The grading on the survey is a “quick reference” guide and the article gives greater details.

Mention has already been made on several occasions of the grading.  I have stated why I consider one to be necessary (yes, the word 'necessary' is purposely used instead of merely 'desirable') but on what are these  gradings going to be based?  If it were possible, then the ideal would be a grading based on the known accuracy of the survey.  However, this is very rarely, if ever, known and therefore some other method must be used.  In many surveys there are closed traverses and it could be assumed that the closure errors on these loops are representative of the whole survey.  The total probable error of the survey could then be determined and the grading based on this figure.  This system would be satisfactory if every survey contained a closed traverse but many do not, especially surveys of individual passages such as those produced of new extensions to caves.  This system cannot therefore be used unless one is going to have one system for those   surveys including a closed traverse and another for those that do not.   This would be most undesirable.

Another basis on which an estimate of accuracy can be based is, on the instruments that are used and the  accuracy with which they are read. The assumption of which this system is based is not ideal because the   accuracy attained with the same instruments used by different surveyors will vary as also will the results by the same surveyor under different caving conditions.  Despite this, a fairly simple system of grading can be devised that will give an indication of the accuracy to be expected.

In 1960, the Cave Research Group of Great Britain published a paper on cave surveying by A.L. Butcher and this included a system of survey grading recommended by the C.R.G. This system is based on the  instruments used, and has been repeated in 'British Caving.'  It is, in my opinion, far from ideal but, as it has been given national publication and is used by practically all cave surveyors it should not be changed now unless anyone can design the perfect answer.  Not agreeing with the Cave Research Group (and many Mendip cavers do not) is not an excuse for trying to replace the present system by one that is only slightly better, if at all.

Having said that the present system is not ideal I should be more explicit and give my reasons for saying this, my main criticism of the C.R.G. gradings is that it appear to have been designed for specific combinations of instruments, and there is no way of arriving at a grading if a different combination is used other than be guessing at the equivalent degree of accuracy.  Partly following from this criticism is my second, that there is no provision for using any form of clinometer to measure slopes until one reaches grade 5.  Figure 1 shows the percentage error that occurs in a plan if the angle of inclination is ignored and it will be seen that a slope of 8o introduces an error of 1% while a slope of 16o gives an error of 4%.  As the angle increases, the error increases even more so, and with an angle of 25° there is a 10% error.  This should show the importance of slope when making a cave survey of any reasonable accuracy.  Even with roughly measured plans of approx. grade 3, it is often useful to take readings of slope, but no credit can be taken for the increased accuracy obtained when using the present C.R.G. System.  If it intended to produce a section as well as a plan, then it is important that angles of inclination should be measured.  Figure 2 shows the changes in height (for various angles of inclination) that are not going to be recorded if no account is taken of slope. For a cave 1,000 feet long, and having an average slope of 10o, a vertical change of 175 feet will be lost. My own experience of cave surveying has shown me that it is extremely difficult to estimate angles in a vertical plane with any accuracy, and this has been borne out by tests on other people. For this reason, it is most desirable to measure angles of slope and not estimate them.  A slope of 5o will not be noticed normally in a cave and while it will introduce an error of half a percent in the plan, three feet will have been lost, in the section with a survey leg of thirty feet.

During the year, the Northern Cavern and Mine Research Society published their own grade of survey. This is just a single grade, approximately equivalent to C.R.G. grade 6, to which they intend making all their surveys. The amazing thing about this grade is, that while they intend to use a tripod mounted prismatic compass marked in half degrees and read to one sixth of a degree, they only consider desirable and not essential, the use of some form of clinometer.  They aim always to measure horizontal distances, to the nearest inch, and then calculate the co-ordinates of their stations using five or seven figure logarithms.  In my opinion, some of these measurements, and the calculations, are going to be considerably more accurate than others and the resulting survey is going to have a far greater error than they intend.

Having made criticisms of the C.R.G. system of grading, can any improvements be suggested?  I made   the criticisms and therefore I will give my suggested modification of the scheme.  My aim is to make the   scheme less specific in the instruments to be used for the survey so that a grading can be obtained, without guesswork, when using a combination of instruments other than one of those listed by Butcher.  It also decreases the guesswork when arriving at a grading after' using an instrument not included in my list,   because the instruments are listed in order of increasing accuracy and because the accuracy of the instruments are sometimes given.

In table 1 will be found a list of instru¬ments, and other means most likely to be used when making a cave survey, and alongside each is given a number.  The scheme simply consists of adding together the numbers shown against each instrument that was used in making the survey, and the result is then that of the cave survey.  One feature of this scheme is that the surveyor is “allowed” to increase or decrease the final grading, thus obtained by half a grade.  This is to take into account factors which cannot easily be written down as hard and fast rules; such factors as the care, with which the surveyor made his instrument readings; the  conditions under which these readings were made; known closure errors, etc. In other words this allows the surveyor latitude to alter the grading slightly either way depending on how accurate he feels the survey should be.

Another feature is the increasing of a grading by half a grade if the 'leap-frog' method is used with hand held instruments.  In this method, the surveyor, instated of starting at Station 1, taking, readings to Station 2,  then moving to 2 and taking readings to 3 etc, starts at Station 2,  takes readings to 1 and 3, then moves to  4 and takes readings to 3  and 5, and so on.

A couple of examples should remove any doubts about the working of the scheme.  Thus, a survey is made using a metal tape and a calibrated compass and a clinometer both mounted on tripods and the   readings being accurate to + or - half a degree, then the numbers are 2 + 4 = 6 and a grading of 5-5 to 6.5 could be claimed.  As another example, if to make a survey a cloth tape; a hand held, prismatic compass and a hand held clinometer (both accurate to + pr - 1o) were used, then the grading would be 1.75 + 1.75 + 1.25 which gives a total of 4.75.  As it is not intended that survey gradings should be given other than as whole or half numbers, then the surveyor would claim either 4.5 or 5 depending on whether or not he thought the survey was as accurate as possible with the instruments used.

It will be found that, in most cases, the gradings obtained with, this scheme agrees with those given against the examples listed on page 393 of “British Caving”.  The main variations occur round the original gradings of 4 and 5.  As already in intimated, the author has always thought that the difference between these   two gradings is very wide; not only does one have to use a calibrated compass to increase from grade 4 to  grade 5, but a clinometer must also be used, and this can lead to a very great increase in accuracy.

Table 2 shows a comparison of gradings between the examples given in “British Caving” and the gradings that are attained by this scheme.  It will be seen that the maximum, grading on this system is 7.5 as compared with a C.R.G. maximum of 7.  However, it is extremely unlikely that anyone making a cave survey with the instruments required for the maximum grading would at the same time be confident enough of his results to claim the extra half grade, knowing what this implies. To keep the results similar to those of the C.R.G., the range of gradings can be limited to any whole or half number between 1 and 7.

A point which arises from a study of table 1 is that normally cave surveys should not be made using instruments that on the table occupy more than two adjacent horizontal lines. If a wider range than this is used, then one of the instruments or methods will be considerably more (or less) accurate than the others.  Any such combination can give rise to a false reading on the scheme.

The only originality, in this scheme is an attempt to standardise a procedure that cave surveyors have been making ever since the Cave Research Group first published their survey gradings - denoting on a cave survey the appropriate grading when a method of survey was not identical with one of the examples they gave.  The variations between the gradings given by this system and those originally described by the C.R.G. are very small, and if this system were adopted there would be no need to alter the gradings. I feel that this system is little, if any, more complicated than the original but is definitely more consistent and more comprehensive.

I have now finished saying my little bit about the grading of cave surveys, but I would like to hear other people’s reactions to my thoughts.  Possibly those of people with experience of surveying would be the most enlightened, but this is not necessarily so as all cavers look at surveys at some time or other.  Now it is your turn.

B.M. Ellis. November   1961.

Table 1

Instruments used to make measurements of….



Inclination and results taken into account when drawing plan.

Estimated out of cave  

Est. noted in cave

Pacing, counting of body lengths.........

Marked string, or cord


Cloth tape



Metal tape, Chain


Tachometer &c













Estimated out of cave  

Est. noted in cave

Hand held compass, readings ± 5o

Hand held prismatic, readings to ± 1o






Estimated out of cave  

Estimated and noted in cave

Hand held clino, readings ± 2o

Hand held clino, readings to ± 1o







If “Leap-Frog” method used with hand held instruments increase above readings by 0.8


Calibrated prismatic compass and clinometer, tripod mounted, readings to ± 0.5o


Theodolite, astrocompass or similar, tripod mounted.  Readings within ± 0,25o of true


Factor available to surveyor to alter final grading to take account of conditions at time of survey, care taken over readings, , known closure errors etc...+ 0.5"


TABLE 1.  Calculation of grading


Table 2.  Comparison of Gradings

Method of Survey

C.R.G. Grade.

New System grade.

Sketch plan from memory, not to scale.



Sketch plan roughly to scale.  No inst. used.  Directions & distances



Simple compass (± 5o) and marked string.



Prismatic compass (± 1o) and cloth tape or marked string.


3-0 - 3-5

Calibrated Pris.  Comp. (± 0.55o) metal tape and clinometer.


4-5 - 5-0

Tripod mounted prismatic compass   (± 0.5o) Clinometer (± 0.55o)



Theodolite, tachometer, metal tape.


6.5 - 7-0

Balch's Hole Extension

by Jill Rollason.

In January 1962, rumour had it that another 1,500 feet of passage, thick with stal, had been discovered in Balch's Hole after entry via a maypole pitch and a trip was arranged for interested B.E.C. members - mainly, photographers.

The extension is a high level passage, entered from Pool Chamber.  It is necessary to climb about fifteen feet on the maypole ladder, and about a further fifteen feet up a steep and difficult rift.  At the head of the rift is a narrow chimney about ten feet deep which leads into the FOURTH CHAMBER, which is richly ornamented with white and cream flowstone, several narrow curtains, and miscellaneous white stalactite.  To  the right can be seen a slope covered with tiny peach-tinted gours and a fine growth of red ''flowers" in a  pool - now dry but with remains of a false floor. The rock appears to be hardly more than compacted clay, and I was glad to move to the next passage which has obviously been shaken in the distant past (possibly by fault movement) but looked a little more reliable.  Here there is a pillar-cum-boss about five feet tall and two and a half foot in diameter which has been cracked into three pieces and moved about a foot out of alignment. The break has not been caused by recent quarry blasting since new stalagmites about four inches tall are growing in the old position.

The  FIFTH CHAMBER, which slopes at about  60 - 70° with a steep boulder scree on the near side, leads to a sump about sixty feet below, and the SIXTH CHAMBER which is angled at about fifty degrees and ends in a bedding plane with two sumps at the bottom. Stalagmite formations are plentiful in both chambers.

I was a little disappointed with this series after the enthusiastic reports which had been given, because I did not think it as attractive as the rest of the cave, but the formations, which have been compared, with  those in September Series in Cuthbert’s, are certainly well worth seeing.

It must be regretfully reported that within these few weeks of the cave’s discovery, many straws have been broken and flowstone ruined by mucky hands - all thoughtlessly and completely unnecessarily.  It is impossible to blame anyone except members of recognised, clubs, since; these are the only people who have been invited to visit the place.

Note 1.  The Maypole has now been replaced by a fixed wood and wire ladder.

Note. 2.  The water filled passage in Pool Chamber described in the previous article has been tested by diving and proved to be merely a pool.


The Belfry Bulletin. Secretary. R.J. Bagshaw, 699, Wells Rd, Knowle ,
Bristol Editor, S.J. Collins, 33, Richmond Terrace, Clifton, Bristol 8.