U.S. Patent No. 2,435,068 – “Number Knitting”
A woman by the name of Virginia Woods Bellamy invented a type of knitting in the 1940s, which she termed “number knitting”. She published a book on the subject in 1952.
When I say “invented”, I mean it in the most legally possible way. You see, Bellamy obtained a U.S. patent for her “number knitting” technique in 1948.
The technique disclosed in her patent is the method commonly known as modular knitting today, in which a larger knitted piece is created from a series of unit pieces which are joined to a previous unit as they are knit. The technique is commonly associated with garter stitch (which is the method taught in Bellamy’s patent), although other stitch patterns may be used.
The basis of Bellamy’s knitting technique was a set of six elementary units: the square, the right triangle, the oblique angled parallelogram, the rectangle, the divided square, and the double parallelogram (Figure 1). The planned design of the knitted piece was laid out on graph paper, where the smallest square on the graph paper grid equalled the elementary square unit. The piece could therefore be measured in terms of the elementary square.
The elementary square was knitted in garter stitch:
[C]ast on X stitches and return (first ridge). This is repeated until X ridges less one ridge have been knitted. On the next row there is binding off with a larger needle. X may equal any number of stitches, usually 2 to 8, or any larger number that is needed for the design, but it must always represent the smallest square used as a unit, i.e., the lowest common denominator.
The patent also noted the use of the elementary square not only as a measure of knitting gauge, but also an estimate of yardage required. A sample elementary square could be knitted, measured, then unravelled and the length of the yarn measured.
The right triangle began with X stitches, and tapered to a point by decreases at the beginning (or end) of every other row. The oblique parallelogram began by casting on X stitches, then increasing at the beginning and decreasing at the end of every other row (or vice versa) until the height of the parallelogram was X or a multiple of X ridges. The rectangle had sides that were a multiple of X (for example, X ridges high and 2X stitches wide). The divided square (known also as a mitered square) was worked on stitches equivalent to two sides of the square (e.g. 2X) with decreases worked at the center of the rows. (That mitered square, by the way, is the type used in the fabled Koigu Oriental Jacket, and the springboard for the fantasic modular creations of Valentina Devine and Horst Schulz, although that doesn’t imply that these designers knew of Bellamy’s work.) Similarly, the double parallelogram was worked on 2X stitches with an increase at the ends and a double decrease at the center of every other row.
As the elementary square had X stitches and X rows, the minimum size of any elementary unit must be X or an integer multiple of X. For example, a double parallelogram may be worked on 4X stitches until 5X garter ridges had been knitted.
With these elementary units, scaled upwards in size where appropriate, a design could be laid out on graph paper and yarn colours selected to highlight any polygonal shape formed from one or more of these units. Figure 2 depicts an afghan featuring some species of animal shape, formed from all the elementary units.
Bellamy also set out rules for number knitting in her patent:
- Knit the units or sections in predetermined order and in accordance with the indications on the diagram, picking up the stitches for a new unit from a preceding one and knitting them as they are picked up. Bind off each unit, except those decreasing in one stitch.
- Turn the diagram, if necessary, to follow the reading of a new unit–the ridges run (one ridge equals two rows) in the same direction as the pencil strokes.
- The number of stitches for each new unit must equal a multiple of the basic square which is the smallest square of the graph paper used for the diagram, and always equals the same number of stitches wide as ridges high.
- Slip the first stitch and purl the last of every row, giving one chain stitch at the side for each ridge.
- Cast on and bind off as loosely [as] possible, or with the needle two sizes larger than the one used for knitting.
- Do not break yarn unless (or until) necessary. When possible, carry from the last stitch–fastened off–of finished unit, to the first stitch of new unit, and cover it by knitting first under then over it.
- Make gauge by casting on the number of stitches given for a basic square, and then knitting the same number of ridges. If basic square is too small to be knitted as gauge (as on one or two or three stitches), make gauge on multiple of basic square.
- Cast off on the wrong side of work unless otherwise directed. Pick up, knitting at the same time, on the right side of the work, unless circumstances dictate otherwise.
The fact that Bellamy obtained a patent does not mean that she somehow owned the entire concept of modular knitting. It’s certainly possible that someone unknown to Bellamy had developed the concept first; or even if Bellamy had known of another progenitor of modular knitting, perhaps Bellamy’s “number knitting” included an inventive enhancement that was not previously known. At any rate, the particular aspects of modular knitting in which Bellamy claimed ownership were set out in eight claims. The first six were directed to a method of making or designing a knitted fabric involving the laying out of a design on graph paper, such as claim 1, below:
1. The method of making a knitted fabric which comprises the laying out on graph paper of a pattern therefor made up of geometric straight-line figures such as basic squares and half squares and pluralities thereof, and dividing the pattern into parts to indicate fabric parts to be knitted in succession; and knitting said parts accordingly, beginning the knitting of each part by knitting it to a preceding part.
The last two claims were directed to a knitted fabric composed of a number of units where the direction of stretch in each piece varied; for example, claim 8:
8. A knitted fabric comprising joint knitted units having greater resistance to stretching along certain lines than the knitted structure as a whole along those lines, and said lines of the units of the knitted structure being arranged to render resistance to stretching substantially the same in a plurality of directions.
That last claim arguably describes a number of knitted pieces with random orientations joined to form a patchwork quilt, which seems a little unfair to makers of crazy quilts from knitted fabrics (if such things were made). However, when a patent application is examined, the resources available to an examiner to determine whether an invention is truly patentable are usually limited to previous publications, typically other patents and published applications, and trade literature brought to the examiner’s attention. The validity of Bellamy’s patent claims could have been challenged in certain proceedings after the patent was issued, but the cost to a would-be infringer in challenging the patent (and even the cost of pursuing an infringer, on Bellamy’s part) might have been excessive compared to the size of the relevant market.
In any event, Bellamy goes down in history as the patent holder for modular knitting, which goes to show that not only is there an advantage to being first, but even to be the first to think, “hey! maybe I could get a patent for this!”
[A brief note: if you need patents and the concepts of originality and inventiveness in the patent context explained, try here (U.S. Patent Office, very brief, and not very informative) or here (U.K. Patent Office, more detail) or here (Canadian Intelletual Property Office, much more detail).]