Einlesen von Zahlen

["U. Jentschura" <ulrich@XXXXXXX.de>]

Liebe DMUG-Nutzer,

bevor ich den unten beschriebenen Sachverhalt als einen Bug-Report an
Wolfram Research schicke, moechte ich fragen, ob ich etwas uebersehen
habe. Weil ich es eventuell als Bug-Report verwenden moechte, sei es mit
gestattet, diesen Sachverhalt auf englisch zu beschreiben:

Viele Gruesse, Uli

===========
Bug Report:
===========

In some calculation I have obtained the following final result:

0.0083727100361487876571543270975126582`22.2208*\[Alpha]^2

Here, \[Alpha]^2 is a power of the fine structure constant (which is
familiar to all physicists). If I load this number into another
Mathematica session, for example by direct copy-and paste, then the
following output results: 

Mathematica 4.0 for IBM RISC System/6000
Copyright 1988-1999 Wolfram Research, Inc.
 -- Motif graphics initialized --

In[1]:= 0.0083727100361487876571543270975126582`22.2208*\[Alpha]^2

                         2
Out[1]= 0.186048 \[Alpha]
=========================

which is not identical to the input number (0.008327 =!= 0.186048).

I assume that the number

0.0083727100361487876571543270975126582`22.2208*\[Alpha]^2

represents, in decimals, a number of 0.0083727100361...

The number is seriously misrepresented when entered into a Mathematica
session via copy-and paste, as above.

My question: Is this a bug, or a feature?

Best Regards,

Ulrich Jentschura