C Misunderstanding Of Variable Assignments.

I have never programmed before, and am starting with some C using tutorials such as howstuffworks.com

In this particular tutorial: HowStuffWorks "How C Programming Works"

A list of Fahrenheit to Celsius conversions are made every 10 degrees F, adding in a line at 98.6 F, then continuing.

I do not understand the given code (the final piece of code on that page) on the loop that adds the 98.6 F line with no bugs.

Code:

 
#include <stdio.h>

int main()
{
    float a, b;
    a = 0;
    b = -1;
    while (a <= 100)
    {
	if ((a > 98.6) && (b < 98.6))
        {
            printf("%6.2f degrees F = %6.2f degrees C\n",
                98.6, (98.6 - 32.0) * 5.0 / 9.0);
        }
        printf("%6.2f degrees F = %6.2f degrees C\n",
            a, (a - 32.0) * 5.0 / 9.0);
        b = a;
        a = a + 10;
    }
    return 0;
}

If b starts at -1, then is attached to a at the end of the loop, and a goes up by 10 each loop, then there is never a point at which a > 98.6 and b< 98.6.

On the end of the 10th loop, a is 90 and goes up by 10 to 100. But starting at -1, b should be going up to 99, which is higher than 98.6. So why does that meet the requirement of a > 98.6 and b< 98.6

I feel that I am misunderstanding the point of b being in the code, because it doesnt seem to matter what b is assigned to in the beginning, as long as it is assigned to a at the end. Where am I going wrong in my thinking?

Thank you so much for your time. I will get better.

b gets the previous value of a; so after the first loop b is 0 and a is 10; at the end of the next loop b is set to the current value of a (10) and then a is increased to 20. And so on.

Oh I see!

b is assigned to a, making b 0, but not till the end of the first loop.
But, why do they assign b to -1 in the beginning? It seems arbitrary, and maybe that's the point.

The program still works with assigning b to any number.

In this case, yes. You can get in trouble for using variables before assigning anything to them; if we didn't assign a value to b before starting, then the first time through the loop b < 98.6 might be true and it might not be true, and it would be an error (logical, not syntax) to try to check it before it had a value.