I've noted in this topic an observation that smoke appeared after the head gasket was changed. May I assume that the carburetor was removed during that gasket change?

In the service information section of this site there is information on the quality & integrity of the carburetor base gaskets, and what might happen if the metering rod vacuum piston pressure source is restricted. (see Fuel Consumption, Excessive carburetor gaskets interchanged SC Vol22 Number 2 page 7). Perhaps that item should be added to the diagnostic list along with the subject of the host of potential difficulties with the automatic choke subsystem.

Steve-52/200
Check the topic "Rear Wheel Bearing Replacement' posted by Packard 1948 . . . lots of shimming information.
Below I present an alternate explanation on what the service technician is striving to accomplish while following the service manual bearing shimming procedure.
Consider the thrust block and both axles, complete with bearings, sitting on your workbench on "V' blocks. Your mental picture should have the splined ends of the axles facing each other with the thrust block between just as they are assembled in the car. Consider the distance between the out-board faces of the bearing cups with zero internal bearing clearance. To that number add the Packard recommended bearing clearance of 0.004 - 0.007 inch. This dimension represents the distance that must be achieved between the inner surfaces of the backing plates of the rear brakes. Here I'm talking about the area where the bearing race touches the backing plate. Shim stock is used to accurately adjust the position of the backing plates. That's the high level view, and mental exercise of what is trying to be achieved.
Now let's consider the bearing cups. The outer diameter of the bearing fits into the 'rear-end' housing bore with a minimum radial clearance, but after sixty to seventy years the fit could have lost a bit of its precision. Since forces can only be transmitted across a bearing at right angles to the roller track, the forces ( that would be the vehicle weight and cornering loads) would tend to push the bearing cup 'hard' onto the backing plate. I doubt whether the cup ever leaves contact with the backing plate during the life of the car, but if it does it's not for long.
About the shims; they are stamped from steel shim stock in 3 different thicknesses, namely 0.005, 0.007, and 0.020 inches. They have 6 total pierced holes, 4 for the bolts that hold the backing plate to the axle housing, 1 for a grease/oil relief passage (so the grease or oil from a leaking seal does not contaminate the friction surfaces of the brake shoes), and finally the large hole in the middle. This center hole is large enough for the bearing race to pass through. The fit of the center hole to the outer diameter of the bearing cup is quite precise on the OEM shims I've encountered. I believe the design intent is for the bearing to touch the face of the backing plate, and the shim will never be contacted by the face of the bearing cup. I equally believe the Packard design engineers expected each rear axle assembly would be selectively shimmed and designed (and exploited) a minimum and maximum expected amount of shims in each unit by controlling manufacturing tolerances of the component parts. By accepting a shimming process during assembly, the tolerances of the individual piece stack-up does not need to be controlled to the 0.003 suggested by the bearing clearance tolerance (wider acceptable tolerances at the piece part level should be easier/lower cost).
I do not believe the procedure presented in the service manual was followed production floor during assembly of the 'rear-end' assembly. One alternate method would be to directly measure the axle detail, and then measure the housing and scribble some arithmetic for the necessary shim stack. However, the most probable alternate method would be to push one bearing cup to a known shim position, which easily could be zero ( no shims using fixtures not backing plates), and then measure how much the other bearing cup is sticking out of the housing. All of the values, including the desired clearance would be combined and the specific shim stack selected, and then divided in two for installation. Once the vehicle has left the production assembly line the service manual procedure is quite adequate, but a bit cumbersome for production. If you have both sides disassembled you could easily get into the right ballpark by using the second suggested alternate procedure.
Initially I was taken by the fact that there are only 3 different shim thicknesses. However by selecting various shims you can find a solution (lots of solutions beyond 0.014) that lies within the tolerance of the 'target' bearing clearance. Try it, quite clever those 'old-timers', and an affordable computer was decades away!
Now on to the 'funny' noises: When the bearing is rolling on the outer race no matter how the backing plate was shimmed the relative position of the backing plate and rotating drum are just as the design engineers expected (assumes nothing is worn-out). Now let's go into a turn and push hard until a bearing reacts out the cornering forces. One bearing is 'taking the load' and all of the brake drum to brake shoe dimensions are OK on that side. On the other side however the tire/wheel/hub/brake drum keeps moving inboard until the axle touches the thrust block and stops only when the thrust block encounters the other axle (which is hard on the bearing). Too much motion of that axle looking for the thrust block can result in the brake drum operating closer to the backing plate (shoes, cylinder, adjuster, etc.) than expected. Extra shims would make the condition worse.
With both rear wheels off the floor, and the brake drag adjusted out, have a 'helper' pull 'their' wheel trying to pull the tire off the car. All the while you're on the other side trying to push your wheel toward the center of the car. While you're pushing and they're pulling try to rotate your side. If the noise is a result of not shimming the bearings correctly you should be able to tell by using this procedure. Reverse your roles (you pull, they push and rotate) . . . the other side is also likely to make noise because the 'shimming' clearance is the same on both sides. With both wheels off the floor and your helper pulling on their tire you should not be able to feel a lot of bearing clearance when you push and pull on your tire, 0.004-0.007 inch is pretty hard to feel. If you're really off on the high side with the amount of shims you'll feel the 'slop'.
Reference the diagram Owen included in his reply. If the noise is a result of excessive 'shimming' I don't know which part might be contacted first but you might see witness marks on the drum's 'diaphragm between the intended friction surface and the hub, (rubbing the metal edge of the shoe, or the side of the friction material into a wear lip in the drum ). I have no idea what the design clearance would be between the stationary parts and the spinning ones inside a drum brake, but we do know the full bearing clearance would need to be accommodated.
I sympathize that modern mechanics don't not study or have experience with historic designs. It would take them some amount of time 'studying' the photos and 'cut-away' drawings to understand the intent.
Hope that helps.
dp

My final thought on the 'unused' hole:
During the manufacture or transportation of the finished fender the hole could have been where a hook would be placed to lift the part . . . especially after the fender was painted.
In the aforementioned Packard history there are several photos of the 'shop floor', which highlight that parts were delivered to the assembly area by an overhead system. I suspect this would be quite typical of the industry. The first photo in the 22nd & 23rd section indicated that the 'front clip' was built as a subassembly. Unfortunately this photo is at the end of the line and does not include detail of the fenders being 'delivered', but at some point the outer fenders would need to join the growing assembly.
In the past I have encountered parts with approximately 3/16 inch edge holes that had no apparent use. I've always assumed that these holes were used to hang the part during final finishing, be it powder coating or spray painting. I guess they'll always be near an edge as are the fender holes.
dp

Howard;
My reference is the photo on page 526 of Automotive Quarterly's 'PACKARD A HISTORY OF THE MOTOR CAR AND THE COMPANY". I've looked at the photo with 9X magnification and have concluded the bumpers were attached with the rolled lip facing up & outboard. In your photo that would be the 'red' cars.
dp

To use household electrical terms the copper binding post is the 'LINE', while the silver post is the 'LOAD'. In automotive terms the 'HOT' wire from the battery (for our 6 volt positive ground cars this would be the negative lead) attaches to the copper post, and the silver is attached to the headlight switch. The stationary post is identified by a copper plating, or a shorter length. The 'Packard' OEM circuit breakers are also marked. The short post is marked 'BAT'. The long post is marked 'AUX'
Internally a circuit breaker is somewhat like a set of points. When all is well the points are in contact, thus providing a circuit. One side of the point set is fixed in the insulated base of the circuit breaker, while the other is mounted on a leaf. It is the leaf that responds to the heat that varies with current. When the circuit breaker 'pops' the leaf defects and opens the circuit. Automotive headlight circuits utilize what is known as a 'Type I', or automatic reset breaker. The leaf motion must be abrupt (snap action) to avoid excessive arcing.
It is a safer design that fixes the 'HOT' lead, and allows the load side to defect. The concern is the failure mode where the leaf defects in such a way that contacts the external case. If the leaf defects into the case, and the circuit breaker was installed correctly there is no short circuit . . . the leaf is not 'HOT'. Just as long as the points re-align as the unit cools this would be known as a 'fault tolerant' design.
The circuit breaker doesn't know about positive versus negative ground circuits, and will work 'OK' with either post attached to the headlight switch, but it is safer to conform to the design intent.
There are European designs that use fuses versus circuit breakers in the headlight circuits. This got me to thinking about what type of failure mode are we trying to protect from that the very first response is to re-energize the circuit . . . and try again . . . and again . . . and again.
The battery is connected to the copper, or, if the posts are not 'color' coded, the shorter post. The brass plate that is part of the light switch is attached to the silver or longer post.
dp

There is also a Packard Hub Cap & Wheel Cover Identification Guide on the website.

One way, but not likely the only way, to access this information is to select:
Packard Literature and Manuals / Year by Year / 1948 / Number 29 in the listing.

Good Luck dp

Try the following process:
1. Start with the parts book in Group 19 and determine from your car's model number whether a 19.041 trim ring is used.
2. Review photos from the Model Information section for the 23rd series and your cars model number.
3. Review photos from the Photo Archive section for the 23rd series ( during the first pass I would stick with factory photos only ).
4. Finally review the photos in both the Photo Archive and Owner Registry.
If the wheel is broken down into two distinct parts, namely the (1) hub and (2) rim, then the various cover covers maybe categorized as:
SMALL - Only center section of the hub is covered
LARGE - The entire hub is essentially covered
TRIM RING - Covers the outer most area of the rim
I believe the higher the trim level of the car, the less the wheel was visible. 'Full Wheel Covers' were not used in the 23rd series. Wheels appear to be body color, but I'll assume the dealer would accommodate 'special requests'. Wire wheels were a few years in the future, again I'll assume a dealer would accommodate an 'upgrade' request from a customer.
dp

and the second file.

dp

Attach file:

---

Size: 1,302.95 KB; Hits: 15

Hi Guys;

I took some photos of my '54 Cavalier, which I believe is 'correct' in the area of the PS pump bracket.

There are two files for a total of 7 photos.

dp

Attach file:

---

Size: 1,485.40 KB; Hits: 16