Pick up live stitches through the FLB to provide the basis for the second stripe. The written instructions and a diagram are at
this link, the photo below shows what this looked like IRL on the blue/green tam.
 |
| The red dotted lines show how the live stitches were picked up through the deco bind-off FLB |
As with the first stripe, slip the first stitch of the stripe color when you come to it to prevent jogging.
On the third round, decrease all the way around by k5, k2tog. I promise you that starting 2/3 of the way around, I fudge the decrease spacing to keep the decreases more-or-less even--no way do I trouble my lazy brain to come out with a correctly divisible number of stitches in the pick-up process, nor has this lazy disinclination to count and figure ever resulted in a noticeably lopsided tam. (All this laziness has to be paid for later, of course, but we will deal with it in good time...)
Work this second stripe the same number of rounds high as the first stripe. Then, just like the first stripe, bind off.
Below is a sketch of where the hat is at this point, as well as a photo showing the hat at this point IRL.
 |
| What the tam theoretically looks like at the end of the second stripe |
 |
| What the tam actually looks like at the end of the second stripe |
Tam Top
As with the previous two stripes, the pick up through the bound-off FLB is worked in the same color as the FLB. Before we go further, however, a little theory.
Smoothing the decrease points--two tricks
The default decrease for a flat-ish top on a hat is 8 stitches decreased every second round. However, if we were to establish 8 decrease points at the beginning of the tam-top and immediately begin decreasing, we would end up with a tam which looked like it were topped with a stop sign--not a good look. We'll use two tricks to minimize this unfortunate look. First, we'll move the decreases away from the very edge of the stripe, which smooths the decrease points out somewhat. Therefore, we'll work two rounds plain without any decreasing, insert our makers on round 3, then work the first decrease as we knit round 4.
Moving the decreases in from the stripe edge smooths out the points, yes, but decreases have a tendency to propagate through the fabric regardless, so more trickery is required. Our second smoothing trick, therefore, is to double the number of decreases to 16 in the first (and only the first!) decrease round, switching to the default 8 decrease rounds thereafter. This works because to the human eye, a 16-sided figure (a hexadecagon) looks a whole lot more like a circle than an 8-sided one does.
However, this leaves us with the somewhat tricky proposition of inserting 16 evenly spaced markers on whatever random number of stitches we happen to wind up with at marker-insertion time. It also requires that we distinguish between the two kinds of decrease when we insert the markers. The below instructions assume you're using a
red marker for the 8 decreases which will continue up the hat, and a
white marker for the 8 intermediate decreases which are worked only once. (Such markers are really easy to make out of a slip-knotted loop of yarn inserted between stitches.)
In theory, we'd now divide the hat top into 8 sections, each with the same stitch count, each headed by red markers. Each of these sections would be further divided by a white marker inserted at the half-way point. The net result would be 16 markers, evenly spaced and inserted alternating red and white.
(Addendum, 11-25-14: thanks to KADRI K. who wrote via e-mail with a correction. If you read this paragraph before the addendum date, the previous text was in error.)
Order out of chaos: marker insert, or "how to evenly divide an uneven number of stitches"
Theory is one thing, but now comes the reality. Except by the most lucky of accidents, you will not have an even multiple of 16 on your needles. We have been knitting in the laziest possible manner, without measuring or counting a darn thing, so this laziness must now be offset with trickiness as we prepare to insert the red and the white markers.
We have several tricks in our arsenal to work with, and which we choose depends on how our stitch count relates to the nearest multiple of 16. There are four possible situations.
- First, your stitch count could be anything between 1 and 7 stitches LOWER than the nearest multiple of 16. If that is the case, you would use TRICK A just after slipping a red marker.
- Next, by luck your stitch count might actually be a perfect multiple of 16. If that is the case, you need not use any tricks, simply insert the stitch markers in an alternating red-white pattern, evenly spaced, however many times around your needle.
- Third, your stitch count might be anything between 1 and 7 stitches HIGHER than the nearest multiple of 16. If that is the case, you would use TRICK B just after slipping a white marker
- Fourth, your stitch count might be exactly half-way in between two multiples of 16. If this is the case, use TRICK C on the white markers.
This is summarized in the below table based on some some common stitch counts you might expect to find in hat-knitting. (Click to enlarge.)
TRICK A:
the FAKE DECREASE trick
Although you wouldn't normally need to do this, it is entirely possible to work a
k2tog (knit two together) followed IMMEDIATELY by a
nearly invisible increase. This looks for all the world as if you had decreased in pattern, although, of course, subtracting a stitch and then immediately adding one back means the stitch count remains the same. This trick works because the decrease shows more than the increase--see for yourself, below is a closeup.
 |
| Fake decrease, closeup |
In this recipe, the fake decrease trick is only to be used in sections headed by the
red markers, because in those locations, a fake decrease will be followed two rounds later by a real decrease, and this actual decrease will propagate through the material. This will eventually create the eye-fooling 16-sided figure despite your fake-decrease trickery two rounds earlier.
Trick A in action
Suppose you have 110 stitches. This is stitch count appears in the FAKE DECREASE column of the table, so that's the trick to do. You're TWO stitches off the perfect multiple. This means that on two different (and separated)
red decrease sections (sections headed by a
red marker) you'll work the fake decrease trick.
See, if you had 112 stitches, you could insert 14 markers 8 stitches apart (7 sections of 16, each of those sections divided into a
red-marker 8-stitch-half and a white-marker 8 stitch-half, for a total of 14 markers). However, since we're two stitches short, and since we're going to work Trick A on the
red markers only, then, as we work round 3 inserting markers, we choose two
red-marker sections at random (not next to one another, however) and on those sections, we'll insert the following
white marker 7 stitches away from the
red marker, rather than the normal 8. We mark these special sections where the stitch count is off by KNOTTING the tail of the
red marker.
At the end of round 3, we have divided our 110 stitches into 12 regular sections--7
white and 5
red--of 8 stitches each, and two special
red-knot sections of 7 stitches each, and all the makers will be inserted in an alternating
red-
white pattern, with the two
red-knot sections not adjoining.
On the actual decrease round (round 4), every time you come to a marker, you slip the marker, work a k2tog. However, on the two
red-knot sections, you'll work a k2tog followed
immediately by an m1. At the end of your decrease round (round 4) you'll have 16 sections, each of which is now 7 stitches wide--12 of those sections will have been decreased from their original 8 stitches by means of the k2tog you worked, and the other two sections--the
red-knot sections--will be 7 stitches wide because the fake decrease trick left them unchanged at their original 7-stitch width. Tada! Trickiness has provided a work-around, and we get an even spacing despite having started with a random stitch count--order has been brought forth from chaos.
TRICK B: the K3TOG trick
It is possible to decrease three stitches into 1 (k3tog) rather than the more usual two stitches into one (k2tog). K3 tog is worked the same exact way as k2tog, just starting one stitch further in. Below is a closeup of the finished result.
 |
| k3tog, closeup |
So, if we count the stitches and find we are a few stitches ABOVE the nearest multiple of 16, we can get rid of these extra stitches via a k3tog instead of a k2tog, spacing the k3togs so they aren't right next to one another, as best as possible. This trick is worked only on the
white (intermediate) markers.
Trick B in action:
Suppose you have 149 stitches. This is stitch count appears in the "K3tog" column of the table, so that's the trick to do. You're FIVE stitches off the perfect multiple. This means that on five different (and separated)
white decrease sections (sections headed by a
white marker) you'll work the k3tog trick.
See, if you had 144 stitches, you could insert 18 markers 8 stitches apart (9 sections of 16, each of those sections divided into a
red-marker 8-stitch-half and a
white-marker 8 stitch-halfhalf, for a total of 18 markers). However, since we're five stitches over, and since we're going to work Trick B on the
white markers only, then, as we work round 3 inserting markers, we choose five
white marker sections at random (as well-spaced from one another as possible) and on those sections, we'll insert the following
red marker 9 stitches away from the
white marker, rather than the normal 8. We mark these special sections where the stitch count is off by knotting the tail of the
white marker.
At the end of round 3, we have divided our 149 stitches into 13 regular sections--4
white and 9
red--of 8 stitches each, and five special
white knot sections of 9 stitches each, and all the makers will be inserted in an alternating
red-
white pattern, with the five
white knot sections as well spaced as possible. Note that with 5 special markers, you can't space the special sections evenly around the brim, but try not to get them all on the same side of the tam.
On round 4 (the actual decrease round) every time you come to a knotted
white marker, you'll slip that marker and work a k3tog, rather than a k2tog. After round 4, you'll have 18 markers, each of which is 7 stitches apart. Again, trickiness has provided a work-around, and we get an order from chaos despite having started with a random stitch count.
TRICK C: the trick to use if you are 8 stitches off a multiple of 16 (halfway between)
If you are halfway between multiples of 16 (8 stitches away from the next higher AND the next lower multiple of 16), insert the
red and the
white markers, but work the first decrease ONLY on the
white markers, then begin the
red marker decreases on the very next round. This decreases away only 8 stitches at the
white makers in the first decrease round, rather than 16, but, because an additional 8 stitches are decreased away at the
red markers only 1 round later, the 16-sided figure is still created, fooling the eye into thinking circle, rather than octagon.
As far as marker-placement for trick 3, you place the
red marker following each
white marker one stitch further apart than the
white marker following a
red marker. This extra stitch in a section headed by a
white marker is decreased away in the first (
white-marker only) decrease round, leaving the 8 red markers all spaced equally apart.
Decreasing after round 4
Regardless of which trick you used, as you knit round 5, remove all the white markers--they marked a once-only decrease location. Also on round 5, you must move the red markers over so that the marker is immediately on the LEFT side of the decrease (whether a real decrease, or a fake decrease) worked earlier. Stated otherwise, unless you move the red markers to the left of the decrease, the spiral created by stacked decreases will not develop properly.
On round 6, and every even-numbered round thereafter, decrease at each red marker by working to with two stitches of the marker, working a k2tog, then slipping the marker. On odd-numbered rounds, knit all the way around with no decreases. Thus, you revert to the default decrease of 8 stitches every second row, which makes a fairly flat hat top.
(Addendum, 11-25-14: Again, many thanks to KADRI K. whose e-mailed correction is now reflected in these decreasing instructions.)
If you are working with a short circular needle, at some point, your stitch count will decrease so far that dpn's, magic loop or two-circs will have to be used.
Alternate ending #1: Kitchener-stitched tailored top
 |
| Kitchener-stitched tailored top |
 |
| Kitchener-stitched tailored top, in real life. Quite elegant. no? |
To make an elegant tailored as on the blue/green tam, the red-marker k2togs continue until there is one stitch between each decrease (16 total stitches).
Pro-tip 3--a truly flat top
If you want a truly flat hat top, switch to smaller needles when there are 5 stitches in each section. This creates a variant on the "truly flat top" for a ribbed cap.
The final 16 stitches are divided onto two dpn's and these two sets of 8 stitches are Kitchener-stitched together, which can be
done with a knitting needle.
Alternate ending #2: I-cord tassels
 |
| I-cord tassels on tam top |
 |
| I-cord tassels in real life |
To make an tassel-top as on the beige/multi tam, the red-marker k2togs continue until the tam top is reduced to two stitches between each decrease (so, 3 stitches for each section, 24 stitches total over the eight sections). Next, the tam is chain-bound-off in the FLB color (beige in the beige/multi tam). As with the previous stripes, the following round is picked up through the chain bind off= 24 stitches. Next, a round was worked in the color you've selected for the I-cord tassels, which is rust in the sample tam. the following round is decreased randomly via k2tog's to 15 stitches. Holding all but three stitches on a scrap yarn stitch holder, the three stitches left live are put onto a dpn and worked into a 3 stitch I-cord of random length. * Again 3 stitches are slipped off the holder onto a dpn and worked into an I-cord of random length. The process is repeated from * until all 15 sts are converted into 5 various-length I-cords. The top of each I-cord is knotted to give each I-cord a cute little ending and add a little heft (this entire process is a variant on
I-cord tassels). A few running stitches were taken between the cords to connect them at the bottom and tame the tendency of the I cords to spread, and the hat is done.
Finishing
Skim in the ends on the inside. You are now ready to block. The entire blocking process is shown in the below photo.
 |
| Blocking a tam in 3 steps-find a plate of the correct size, insert the plate and soak, stand to dry on a pedestal made from a tall glass, remove plate when tam is dry |
Begin by finding a plate (mint green in illustration) in your desired tam-top size.
Pro-tip 4--blocking forms
For a tam, a full-size dinner plate usually does the trick, but if not, thrift stores are a great resource for cheap-o plates in all sizes. Or, cut a blocking form from a flexible cutting board--a pack of these are handy for cutting out blocking forms for many small objects: tam-blockers as well as sock- mitten- and even glove-blockers.
The plate can be somewhat larger than the dry tam top, or the same size, but not smaller. Find also, a tall glass, stable enough to support the plate (red in illustration). Fit the plate, convex (bottom) side up, inside the dry tam, then wet the tam-plate sandwich by submerging in a sink partly filled with lukewarm water, as shown.
Once the tam is utterly soaking (a few minutes, at least) remove the tam from the water, turn the works upside down, then place the glass under the plate so the tam is suspended with even its bottom band clear of the underlying surface. Pat and smooth the tam over the plate, stretching it to fit evenly, if necessary.
 |
The FLB between stripes 1 and 2 is the widest part of the tam, block accordingly |
The FLB between stripes 1 and 2 is the largest part of the tam, so that's what runs around the rim of the plate.
The tam will now drip dry on its pedestal. Annoyingly, this takes a day at least, since the plate slows the drying time. It is best to leave the tam to dry over a thick and thirsty towel on a waterproof surface--a tile floor or kitchen counter. When utterly dry, remove the plate and the tam will be smoothly blocked.
If you choose to add a
pom-pom, these are attached after the blocking process is over.
 |
| Pom pom added to tailored (Kitchener-stitched) top |
 |
| Pom pom in real life |
Pro-tip 5--pom poms and washing
Pom-poms on hats are best removed before washing--a washed pom-pom looks awful. Therefore, sew your pom pom down using a different color yarn-it only takes a few big stitches to tack down--and this makes it easy to snip free for washing. Or, you know, use a safety pin. Lazy, yes. The kind for emergency button repair, with a bump halfway down are best, just be sure the metal of the pin penetrates the solid heart of the pom-pom.
Good knitting