Bergdala Spinnhus


To read older weaving literature

Older weaving literature often uses a language that we do not readily understand – or at least do not understand precisely.

This article is concerned with "Praktisk vävbok tillägnad den idoga Svenska kvinnan" by Nina v Engeström (1st ed 1896) – now available in an English translation from Unicorn books, or from the publisher Bokförlaget Rediviva.

About yarn packages and numbers

On page 5 the first somewhat obscure table can be found: (my translation)

"       About yarn and reeds
A bundle of cotton yarn no 8 usually contains 8 hanks
[...]
One hank cotton yarn is about 720 meters = 1200 alnar (note 1) and will suffice for about 12 meters = 20 alnar, 1 lea (note 2) wide. For example, if one is to wind a warp 12 meters 30 leas, bobbin 30 hanks."

The 720 meters give us an idea that the yarn numbering system used in 1896 is the same as is used today, the cotton count (also abbreviated Ne) (note 3).

But how much is a lea? Calculating backwards from the 12-meter warp above, the "lea" will result in 60 ends (720/12 = 60). This figure can also be found in "Ny väfbok II" (see appendix: excerpt in Swedish) – but there it is also said that the number of ends in a "lea" differs between parts of the country.  

About reed sizes

On page 6 there is a table of reeds that also looks unfamiliar (my translation):

"  Information about the, in ordinary circumstances, used reeds for different counts of cotton yarn:

For number  8 ........ 120 to 130 dents in 15 cemtimeters = 1 qvarter (note 4)
  "      "       10 ........ 130 to 140
[...]
  "      "       26 ........ 200 …… 2 ends in each dent.  "

First, it can be seen that "2 ends in each dent" is probably meant to be applied to the whole table. 120 dents in 15 cm corresponds to 8 dents/cm (or 20 dents/inch). Experience tells that that number is far too low for a hard-wearing fabric – the doubled number would be more appropriate. (Remember that a singles cotton no 8 has approximately the same grist as has a two-ply cotton no 16, also written 16/2.)

Next question what the reed numbers mean, compared to modern reeds. And why are they given in 15 cm = 1 qvarter?

A reasonable explanation, even if only built on guesswork, can be the following:

Before 1878, when the SI system (the decimal system for weights and measures, with kilogram, meter and second as the base units) was, by law, introduced in Sweden, the foot was used as the base length unit. One Swedish foot was 0,2969 meters, and was divided into 12 inches ("verktum"). 1 inch was divided by successive halving: 1/2, 1/4, 1/8, 1/16, 1/32, 1/64 etc. (Note that one imperial foot is 0,3048 meters, so the Swedish inch was slightly shorter.)

Reeds for weaving are traditionally manufactured by wrapping a thread or a cord around the wooden supports, then one spline is inserted, the cord is wrapped, next spline is inserted and so on, until the reed is complete.

Let us assume that the cords used for the distances were made in different (standard) grists, and were defined by their diameter. This would mean that the divisions between dents would come to be – by standard - 1/2, 1/4, 1/8, 1/16, 1/32, 1/64 (and so on) inch. Added to these measurements there were probably used 1/12 and multiples thereof, as 12 is a common base for inch calculations.

As a base for the reed numbers it looks as if the number of dents per half foot, ie the number of dents per 6 inches, was used. This way of giving a measure can be compared to how the measure of nets and woven filters and sieve material in some countries still is given in "mesh" (number of "holes" per inch), instead of the width of each "hole" itself.

Let us see where we get with this reasoning:

Division number of dents/6 inches      
1/2 12
1/4 24
1/8 48
1/12 72
1/16 96
1/24 144
1/32 192
1/48 288
1/64 384
In the middle of the 19th century, a parallell division of the inch was introduced (probably as a preparation for the decimal system, already decided on), namely a decimal division. This meant the inch was divided into 10. Let us (hypothetically) introduce a division into 1/10, 1/20, 1/30 and 1/40 inches in the table:
Old division Decimal division Number of dents/6 inches  
1/2 12
1/4 24
1/8 48
1/10 60
1/12 72
1/16 96
1/20 120
1/24 144
1/30 180
1/32 192
1/48 288
1/64 384
The steps between the different reed numbers get, with this division, too big, so it is reasonable to suppose that in-between steps were introduced:
Old division Decimal division "in-between" Number of dents/6 inches
1/2 12
1/4 24
1/6 36
1/8 48
1/9 54
1/10 60
1/11 66
1/12 72
1/14 84
1/16 96
1/18 108
1/20 120
1/22 136
1/24 144
1/28 168
1/30 180
1/32 192
1/48 288
1/64 384

 
We can now note that most of the reed densities referred to in the text can be explained by the manufacturing method, with standard-sized cords given in the old inch-system.
Some of the reeds mentioned in the text do not conform to any of the above figures, for example 75, 82 or 57 dents/15 cm. These diversions are easily explained if we take into consideration the precision of the cords, and the precision of the work of the reed-maker. Some diversion from the standard measures (the number of dents per 6 inches) given in the table above must be seen as reasonable. This means that a reed that in reality has, for example, 57 dents per 6 inches probably was meant to have a division of 1/9 inch per dent, or possibly 1/10 inch/dent – which would correspont to either a 54-dent or a 60-dent reed. Which of those it was meant to be can not be guessed at.

The metric system was introduced (by law) in Sweden in 1878. It is reasonable to suppose that the old divisions with 12, 24, 36, 48, 54, 60, 66, 72, 84, 96 ... dents per 6 inches, instead went to 10, 20, 30, 40 ... dents per 6 inches. This way, the reed numbering got easier, while still resembling the old numbering system.

At the same time the 15 cm unit was substituted for the 6 inch unit. The old series was adapted to the new measurements: 24 and 36 dents/6 inches were changed to 30 dents/15 cm, 36 dents/6 inches were changed to 40/15 cm, 48/6 inches became 50/15, 60 remained, 72 became 70 ...

Today Swedish reed numbers are normally given in number of dents per 10 cm, and the divisions are usually in 5 dent increments: 20/10, 25/10, 30/10, 35/10 and so on.

Re-calculating the table on page 6 in the book, adapting the results to modern reeds, we get:

120/15 80/10 20 dents per inch  
130/15 85/10 (actually 86)
140/15 90 or 95/10 (actually 93)
150/15 100/10 25
160/15 105/10 (actually 106) 28 (actually 27)
170/15 110/10 (actually 113) 28 (actually 29)
180/15 120/10
200/15 130 or 140/10 (actually 133)
 
In countries where the foot-based system is still in use, the reed numbers are given in number of dents per inch. The right column above gives the common inch-based reed numbers.  

Note 1

1 aln = 2 swedish feet = 0,5937 m = 23,4 inches (modern)
back  

Note 2

Swedish pasma. A lea is a number of ends in a yarn package, and varies between regions, countries and types of yarn. Strictly speaking every "lea" has to be defined in its own context.
back  

Note 3

"Cotton Spinning: its development, principles and practice",
Richard Marsden, Meb. Soc. Arts, Editor of "The Textile Manufacturer"
Published in London by George Bell and Sons, 3rd edition, 1888, page 329:
Counts of yarn. - The denomination of yarns in this and other countries where the English system prevails is obtained from the number of hanks of 840 yards each in the 1 lb Troy. Thus, of 20s. there are twenty hanks, or 30s. thirty hanks in 1 lb and similarly with other counts.

Cotton Yarn Measure.
54 inches = 1 thread (or circumference of wrap reel)
4320 inches = 80 threads = 1 lea
30,240 inches = 560 threads = 7 leas = 1 hank, or 840 yards.

back

 

Note 4

1 qvarter = 6 Swedish inches = 0,1484 m = 5,8 inches (modern)
back  

Appendix: excerpt from "Ny väfbok II", publ by Östergötlands Läns Hemslöjdsförening, printed 1913

Numrering och beräkning af Bomullsgarner.
  För bomullsgarner användes engelsk numrering, där numret anger antalet pasmor à 840 "yards" som går på ett eng. lb. (1 yard = 0,914 m., 1 eng. lb = 0,453 kg).
  Vid fabrikerna förpackas bomullsgarnerna i bundtar, indelade i härfvor, pasmor och hank sålunda:
  En bundt enkelt garn innehåller 20 härfvor (= vred) = 10 eng. lb;
  Ett eng. lb = 2 härfvor innehåller så många pasmor som garnet har N:o.
  En pasma innehåller omkring 760 m. fördelade på 7 hank (= knäpp).
  Ex. N:o 10 har 10 pasmor på hvarje eng. lb = 100 p. pr bundt. N:o 15 har 15 p. pr lb = 150 p. pr bundt o. s. v. För medelfina tvinnade blekta garner beräknas garnets förkortning genom tvinningen till 5 %. För hårdare tvinnat garn beräknas omkring 10 %. En pasmas medellängd skulle då blifva 700 à 720 m. För oblekta garner något mindre.
  Sättet att beräkna garnåtgången till en väf äro mycket olika i olika provinser. I södra Sverige indelas varpens trådantal i "pasmor" eller "tal" = 60 tr., hvilket äfven är förhållandet i norra delen af vårt land. Följande uppgifter hafva lämnats från Rodenstamska skolan, Hudiksvall: "en pasma tvinnat bomullsgarn räcker till en m. lång väf, 10 pasmor bred (en pasma = 60 tr.) eller 10 m. lång väf, en pasma bred."
  I mellersta Sverige indelar man varptrådarne i "bund". Ett bund = 100 tr. "Så många pasmor, så många bund till hvar elfte aln", säger man i Västergötland och Småland.

back

Illustration: a "bundle" of old cotton yarn had 20... hanks?. Each hank held 7 pieces of 60 turns, each of them approximately 720 m. The yarn number is not given, but someone has written 20/2 on the paper. - My package therefore only conforms partly to the description above. Either "hank" or "lea" or "knäpp" is not there...

an old "bundle"

held 20 hanks

of 7 lea, each 760 m
  © Kerstin Fröberg 2009