* * *

I didn't make up a pattern for this one, because I was looking forward to working on the fly. I knew the approximate width that I wanted, so I made up a quick stitch sample to figure out the calculations for how many stitches were required and what the base chain count would be.

After that point, it was all about playing with colour and the combination of stitches.

One thing I did find out: when using different stitches, it's very easy to veer off from producing a clean, straight vertical edge as you go along. I dropped stitches somewhere along the way. It all seemed to come down to how the rows were ended and started when turning. To keep everything even, I opted to use the five double crochet shell stitch as my baseline. For my blanket, I had forty-two complete clusters per row. I made sure that I had at least one of these rows for every ten rows of crochet. If I was a few stitches under or over on a row I worked on, I could easily shift the shell cluster(s) one way or the other. Besides, it's easier to count forty-two stitch clusters than two hundred-something, individual stitches.

Because these clusters are separated by two blank stitches on either side of their centre point, I was able to increase or decrease without having to pull apart previous rows. Checking on the total stitch count this way kept my edge even without having to count every row and every stitch made.



Once I completed thirty-seven clusters, I marked it with a safety-pin and continued on to complete the row. If any adjustments were required, I pulled out my work, to the point of the thirty-seventh cluster. In my case, I always had to increase stitches, so some of my clusters were divided by a single stitch on one side instead of the normal two stitch separation on both sides.

If you need to decrease stitches, increase the two stitch separations up to three, for one side of the cluster(s). The important thing to remember is to end up with the total number of five-stitch clusters that you started out with. This will keep your stitch count relatively accurate as you keep adding rows.

The stitch clusters are forgiving in how they interlock with each other and varying stitch groupings, from row to row. Corrections made within these types of stitch rows were in no way visible. I also took care not to make the adjustment right at the edge, but somewhere between the thirty-seventh cluster and the end of the row.

I think it's apparent that I can be a little fanatical about keeping my edges straight.





The border is made up of fan stitch clusters and one outer row of half-double crochet. I like the results of this experiment so much, that there's no way that I am pulling this one apart. EVER.